Dissertations / Theses on the topic 'Batch optimization'

Specialty chemicals industry relies on batch manufacturing, since it requires the frequent adaptation of production systems to market fluctuations. To be first in the market, batch industry requires decision-support systems for the rapid development and implementation of chemical processes. Moreover, the processes should be competitive to ensure their long-term viability. General-purpose and flexible plants and the consideration of physicochemical insights to define an efficient operation are also cornerstones for the success of specialty chemical industries. Precisely, this thesis tackles the systematic development of batch processes that are efficient, economically competitive, and environmentally friendly, to assist their agile introduction into production systems in grassroots and retrofit scenarios. Synthesis of conceptual processing schemes and plant allocation subproblems are solved simultaneously, taking into account the plant design. With this purpose, an optimization-based approach is proposed, where all structural alternatives are represented in a State-Equipment Network (SEN) superstructure, following formulated into a Mixed-Logic Dynamic Optimization (MLDO) problem which is later solved to minimize an objective function. Essentially, the strength of the proposed methodology lies in the modeling strategy which combines the different kinds of decisions of the integrated problem in a unique optimization model. Accordingly, it considers: (i) synthesis and allocation alternatives combination, (ii) dynamic process performance models and dynamic control variable profiles, (iii) discrete events associated to transitions of batch phases and operations, (iv) quantitative and qualitative information, (v) material transference synchronization to ensure batch integrity between unit procedures, and (vi) batch and semicontinuous processing elements. Different strategies can be used to solve the resulting MLDO problem. A deterministic direct-simultaneous approach is first proposed. The mixed-logic problem is reformulated into a mixed-integer one, which is fully-discretized to provide a Mixed-Integer Non-Linear Programming (MINLP) that is optimized using conventional solvers. Then, a Differential Genetic Algorithm (DGA) and a hybrid approach are presented. The purpose of these evolutionary strategies is to pose solution alternatives that keep solution goodness while seek for the improvement of computational efficiency to handle industrial-size problems. The optimization-based approach is applied in retrofit scenarios to solve the simultaneous process synthesis and plant allocation, taking into account the physical restrictions of existing plant elements. The production of specialty chemicals based on a competitive reactions system in an existing reactor network is first defined through process development and improvement according to different economic scenarios, decision criteria, and plant modifications. Additionally, a photo-Fenton process is optimized to eliminate an emergent wastewater pollutant in a given pilot plant, pursuing the minimization of processing time and cost. Batch process development in grassroots scenarios is also proven to be a problem of utmost importance to deal with uncertainty in future markets. Seeking for plant flexibility in several demand scenarios, the expected profit is maximized through a two-stage stochastic formulation that includes simultaneous plant design, process synthesis, and plant allocation decisions. A heuristic solution algorithm is used to handle the problem complexity. A grassroots plant design is defined to implement the previous competitive reaction system, where decisions like the feed-forward trajectories or operating modes allow the adaptation of master recipes to different demands. Finally, an acrylic fiber production example is presented to illustrate process development decisions like the selection of tasks, technological alternatives, chemicals, and solvent reuse.
La indústria de productes químics especials es basa en la fabricació discontinua, ja que permet adaptar de forma freqüent els sistemes de producció en funció de les fluctuacions de mercat. Per ser líder al sector, són necessàries eines de suport a la decisió que ajudin a l’àgil desenvolupament i implementació de nous processos. A més, aquests han de ser competitius per garantir la seva viabilitat a llarg termini. Altres peces clau per una operació eficient són l’ús de plantes flexibles així com l’estudi dels fenòmens fisicoquímics. Aquesta tesis aborda justament el desenvolupament sistemàtic de processos químics discontinus que siguin eficients, econòmicament competitius i ecològics, per contribuir a la seva ràpida introducció en els sistemes de producció, tant en escenaris de plantes existents com des de les bases. En concret, es planteja la resolució simultània de la síntesi conceptual d’esquemes de procés i l’assignació d’equips, tenint en compte el disseny de la planta. Amb aquest objectiu, es proposa una metodologia de solució basada en optimització, on les alternatives estructurals es representen en una Xarxa d’Estats i Equips (SEN per les sigles en anglès) que es formula mitjançant un problema d’Optimització Dinàmica Mixta-Lògica (MLDO per les sigles en anglès) que es resol minimitzant una funció objectiu. La solidesa de la metodologia proposada rau en la estratègia de modelat del problema MLDO, que integra els diferents tipus de decisions en un sol model d’optimització. En concret, es consideren: (i) la combinació d’alternatives de síntesi i assignació d’equips, (ii) models de procés i trajectòries de control dinàmics, (iii) esdeveniments discrets associats al canvi de fase i operació, (iv) informació quantitativa i qualitativa, (v) sincronització de transferències de material en tasques consecutives, i (vi) elements de processat discontinus i semi-continus. Existeixen diverses estratègies per resoldre el problema MLDO resultant. En aquesta tesi es proposa en primer lloc un mètode determinístic directe-simultani, on el model mixt-lògic es transforma en un mixt-enter. Aquest es discretitza al seu torn de forma completa per obtenir un problema de Programació No-Lineal Mixta-Entera (MINLP per les sigles en anglès) el qual es pot resoldre utilitzant algoritmes d’optimització convencionals. A més, es presenten un Algoritme Genètic Diferencial (DGA per les sigles en anglès) i un mètode híbrid. Totes dues estratègies esdevenen alternatives de cerca amb l’objectiu de mantenir la bondat de la solució i millorar l’eficàcia de computació per tractar problemes de dimensió industrial. La metodologia de solució proposada s’aplica al desenvolupament de processos discontinus en escenaris de plantes existents, tenint en compte les restriccions físiques dels equips. Un primer exemple aborda la manufactura de productes químics basada en un sistema de reaccions competitives. Concretament, es desenvolupa i millora el procés de producció implementat en una xarxa de reactors considerant diferents escenaris econòmics, criteris de decisió, i modificacions de planta. En un segon exemple, s’optimitza el procés foto-Fenton per ser executat en una planta pilot per eliminar contaminants emergents. Buscant integrar el desenvolupament de procés i el disseny de plantes flexibles en escenaris de base, es presenta una formulació estocàstica en dues etapes per a optimitzar el benefici esperat d’acord a diversos escenaris de demanda. Per gestionar la complexitat d’aquest problema es proposa la utilització d’una heurística. Com a exemple, es planteja el disseny d’una planta de base on implementar l’anterior sistema de reaccions competitives. Decisions com les trajectòries dinàmiques de control o la configuració d’equips permeten adaptar la recepta màster en funció de la demanda. Un darrer exemple defineix el procés de producció de fibra acrílica, il·lustrant decisions com la selecció de tasques, tecnologia, reactius o reutilització de dissolvents.
La industria productos químicos especiales se basa en la fabricación discontinua, la cual permite la adaptación frecuente de los sistemas de producción en función de las fluctuaciones de mercado. Para ser líder en el sector, son necesarias herramientas de soporte a la decisión que contribuyan al ágil desarrollo e implementación de nuevos procesos. Además, éstos deben ser competitivos para garantizar su viabilidad a largo plazo. Otras piezas clave para una operación eficiente son la utilización de plantas flexibles y el estudio de los fenómenos fisicoquímicos. Esta tesis aborda justamente el desarrollo sistemático de procesos químicos discontinuos que sean eficientes, económicamente competitivos y ecológicos, para contribuir a su rápida introducción en los sistemas de producción, ya sea en escenarios de plantas existentes o desde las bases. En particular, se plantea la resoluciónsimultánea de la síntesis conceptual de esquemas de proceso y la asignación de equipos, teniendo en cuenta además el diseño de planta.Con este fin, se propone una metodología de solución basada en optimización, donde todas las alternativas estructurales se representan en una Red de Estados y Equipos (SENpor sus siglas en inglés) que se formula mediante un problema de Optimización Dinámica Mixta-Lógica (MLDO por sus siglas en inglés) que se resuelve minimizando una función objetivo. La solidez de la metodología propuesta reside en la estrategia de modelado delproblema MLDO, que integra los diferentes tipos de decisiones en un solo modelo de optimización. En concreto, se consideran: (i) la combinación de alternativas de síntesis y asignación de equipos, (ii) modelos de proceso y trayectorias de control dinámicos, (iii)eventos discretos asociados al cambio de fase y operación, (iv) información cuantitativa y cualitativa, (v) sincronización de la transferencia de material en tareas consecutivas, y(vi) elementos de procesado discontinuos y semicontinuos.Existen diversas estrategias para resolver el problema MLDO resultante. En esta tesis se propone en primer lugar un método determinístico directo-simultáneo, donde el problema mixto-lógico se reformula en un mixto-entero. A su vez, éste se discretiza de formacompleta para obtener un problema de Programación No-Lineal Mixta-Entera (MINLP por sus siglas en inglés) el cual se puede resolver mediante algoritmos de optimización convencionales. Además, se presentan un Algoritmo Genético Diferencial (DGA por sussiglas en inglés) y un método híbrido. Ambas estrategias se plantean como alternativas de búsqueda con objeto de mantener la bondad de la solución y mejorar la eficacia de computación para tratar problemas de dimensión industrial.La metodología de solución propuesta se aplica al desarrollo de procesos discontinuos en escenarios con plantas existentes, teniendo en cuenta las restricciones físicas de los equipos. Un primer ejemplo aborda la fabricación de productos químicos basada en un sistema de reacciones competitivas. En concreto, se desarrolla y mejora el proceso de producción a implementar en una red de reactores considerando diferentes escenarios económicos, criterios de decisión, y modificaciones de planta. En un segundo ejemplo,se optimiza el proceso foto-Fenton a ser ejecutado en una planta piloto para eliminar contaminantes emergentes.Persiguiendo la integración del desarrollo de proceso con el diseño de plantas flexi-bles en escenarios base, se presenta asimismo una formulación estocástica en dos etapas para optimizar el beneficio esperado de acuerdo a varios escenarios de demanda. Paramanejar la complejidad de dicho problema se propone la utilización de una heurística.Como ejemplo, se plantea el diseño de una planta de base para implementar el anterior sistema de reacciones competitivas, donde decisiones como las trayectorias dinámicas de control o la configuración de equipos permiten adaptar la receta máster en función de lademandas. Por último, se presenta un ejemplo donde se define el proceso de producción de fibra acrílica, ilustrando decisiones como la selección de tareas, alternativas tecnológicas, reactivos químicos o la reutilización de disolventes.

Multivariate statistical process control (MSPC) techniques play an important role in industrial batch process monitoring and control. This research illustrates the capabilities and limitations of existing MSPC technologies, with a particular focus on partial least squares (PLS).In modern industry, batch processes often operate over relatively large spaces, with many chemical and physical systems displaying nonlinear performance. However, the linear PLS model cannot predict nonlinear systems, and hence non-linear extensions to PLS may be required. The nonlinear PLS model can be divided into Type I and Type II nonlinear PLS models. In the Type I Nonlinear PLS method, the observed variables are appended with nonlinear transformations. In contrast to the Type I nonlinear PLS method, the Type II nonlinear PLS method assumes a nonlinear relationship within the latent variable structure of the model. Type I and Type II nonlinear multi-way PLS (MPLS) models were applied to predict the endpoint value of the product in a benchmark simulation of a penicillin batch fermentation process. By analysing and comparing linear MPLS, and Type I and Type II nonlinear MPLS models, the advantages and limitations of these methods were identified and summarized. Due to the limitations of Type I and II nonlinear PLS models, in this study, Neural Network PLS (NNPLS) was proposed and applied to predict the final product quality in the batch process. The application of the NNPLS method is presented with comparison to the linear PLS method, and to the Type I and Type II nonlinear PLS methods. Multi-way NNPLS was found to produce the most accurate results, having the added advantage that no a-priori information regarding the order of the dynamics was required. The NNPLS model was also able to identify nonlinear system dynamics in the batch process. Finally, NNPLS was applied to build the controller and the NNPLS method was combined with the endpoint control algorithm. The proposed controller was able to be used to keep the endpoint value of penicillin and biomass concentration at a set-point.

Dynamic modelling and optimization of three different processes namely (a) bulk polymerization of styrene, (b) solution polymerization of methyl methacrylate (MMA) and (c) emulsion copolymerization of Styrene and MMA in batch and semi-batch reactors are the focus of this work. In this work, models are presented as sets of differential-algebraic equations describing the process. Different optimization problems such as (a) maximum conversion (Xn), (b) maximum number average molecular weight (Mn) and (c) minimum time to achieve the desired polymer molecular properties (defined as pre-specified values of monomer conversion and number average molecular weight) are formulated. Reactor temperature, jacket temperature, initial initiator concentration, monomer feed rate, initiator feed rate and surfactant feed rate are used as optimization variables in the optimization formulations. The dynamic optimization problems were converted into nonlinear programming problem using the CVP techniques which were solved using efficient SQP (Successive Quadratic Programming) method available within the gPROMS (general PROcess Modelling System) software. The process model used for bulk polystyrene polymerization in batch reactors, using 2, 2 azobisisobutyronitrile catalyst (AIBN) as initiator was improved by including the gel and glass effects. The results obtained from this work when compared with the previous study by other researcher which disregarded the gel and glass effect in their study which show that the batch time operation are significantly reduced while the amount of the initial initiator concentration required increases. Also, the termination rate constant decreases as the concentration of the mixture increases, resulting rapid monomer conversion. The process model used for solution polymerization of methyl methacrylate (MMA) in batch reactors, using AIBN as the initiator and Toluene as the solvent was improved by including the free volume theory to calculate the initiator efficiency, f. The effects of different f was examined and compared with previous work which used a constant value of f 0.53. The results of these studies show that initiator efficiency, f is not constant but decreases with the increase of monomer conversion along the process. The determination of optimal control trajectories for emulsion copolymerization of Styrene and MMA with the objective of maximizing the number average molecular weight (Mn) and overall conversion (Xn) were carried out in batch and semi-batch reactors. The initiator used in this work is Persulfate K2S2O8 and the surfactant is Sodium Dodecyl Sulfate (SDS). Reduction of the pre-batch time increases the Mn but decreases the conversion (Xn). The sooner the addition of monomer into the reactor, the earlier the growth of the polymer chain leading to higher Mn. Besides that, Mn also can be increased by decreasing the initial initiator concentration (Ci0). Less oligomeric radicals will be produced with low Ci0, leading to reduced polymerization loci thus lowering the overall conversion. On the other hand, increases of reaction temperature (Tr) will decrease the Mn since transfer coefficient is increased at higher Tr leading to increase of the monomeric radicals resulting in an increase in termination reaction.

The injection stretch blow Moulding (ISBM) process is widely used to manufacture PET bottles for the beverage and consumer goods industry. However, at present the ISBM process, including both the stretch blow moulding and reheating processes, are totally open-loop. The processes are complex and users often have to rely heavily on trial and error method to set up and control it. In this research, with the main purpose to displace this experience-based trial and error method in the ISBM process, the associated modelling and control methods are proposed.

Nowadays, minimization of the negative environmental impact of manufacturing processes is considered one of the most challenging problems in various industrial fields. Research communities and environmental legislators are continuously working to address these problems by placing significant efforts in devising new strategies to increase environmental sustainability. One of these problems is the lack of a comprehensive framework that can simultaneously improve economic aspects and lessen the impact on the environment. The need for a mathematical model that can assist firms in reaching suitable investment decisions has become of paramount importance. In this context, this study aims at optimizing the environmental and economic sustainability of batch production systems (i.e. a series of workstations where products are manufactured in batches). To this end, a profit maximization model was created by incorporating constraints such as budget, demand, greenhouse gas emissions and hazardous wastes within the manufacturing stage of product life cycle. Moreover, the model provides detailed guidelines on required improvements in a specific manufacturing system and calculates the investment associated with such implementations. This new approach was tested by using two different software packages and results were probed and discussed in different scenarios to investigate its validity. Sensitivity analysis and simulation results proved the consistency of the proposed mathematical model. In particular, in order to further assess the validity of the model, a pharmaceutical plant was selected as a case study, which also permitted discussion on additional aspects of the problem.

F-18 is a short-lived radioisotope commonly used in Positron Emission Tomography (PET). This radionuclide is typically produced through the O-18(p,n)F-18 reaction by proton bombardment of O-18-enriched water. Thermosyphon batch targets have been proposed as a means to increase F-18 production due to their enhanced heat rejection capabilities. These boiling targets have been operated with up to 3.2 kW of beam power with manageable O-18 enriched water volumes. The purpose of this research project has been to develop computational methods which can be used to design new targets with enhanced production capabilities. The computational methods developed in this work were used to design a low power thermosyphon production target for the Duke Medical Center cyclotron. This design was modeled to be range thick, and operate within the desired margins for beam powers in excess of 1 kW, the operating limit of the Duke cyclotron. A sensitivity analysis of the computational methods was performed which indicated the model is most sensitive to the boiling and condensing heat transfer coefficients. Even with a high uncertainty in these coefficients, the target should still operate well within the desired margins.

Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, 2019, In conjunction with the Leaders for Global Operations Program at MIT
Thesis: S.M., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2019, In conjunction with the Leaders for Global Operations Program at MIT
Cataloged from PDF version of thesis.
Includes bibliographical references (page 55).
J&J Vision Care (JJVC) uses production scheduling methods that are not fully optimized, causing over-production of certain SKUs, and reducing capacity for other SKUs on backorder. This makes planning a weekly run-schedule for each line difficult. It is also difficult to understand where to invest capital to create an optimally flexible fleet of production lines. JJVC is currently capacity-constrained, so optimizing the production to increase output will directly translate to additional revenue. The three main areas that the leadership team wants to explore in this project are: 1. What is our current fleet flexibility? 2. How much capacity can be freed up if our fleet was more flexible? 3. Can we create a cost modeling tool that will provide more granularity in brand and sales channel profitability? First, the brands and SKUs on each line that are "validated" to run (by FDA, etc.) must be quantified.
Not all validated SKUs on a line are "runnable" though: Process issues often arise in the plant that prevent some of these validated SKUs from being produced (e.g. mechanical tolerances, chemistry, etc.). Therefore, the gap between validated and runnable SKUs will be an opportunity to explore. One constraint originally studied was the "runnable" vs "validated" prescriptions at the Jacksonville site; The percentage of runnable vs validated SKUs is only 73%, meaning that 27% of the prescriptions that J&J invested time and money to validate cannot be produced on certain lines due to manufacturing issues. The impact of this constraint and others can be quantified to identify improvement opportunities. Second, potential additional capacity can be calculated by running a sensitivity analysis with the planning tool (i.e. the optimization model) to analyze how outputs (e.g. throughput, changeover times, etc.) are affected by changing certain inputs: Mold, core, and pack change times, production rate, minimum lot sizes, service level, etc. It is also possible to change the objective function to place more weight on certain user-defined parameters.
The impact of these changes were observed by collecting the master planning data for a defined time-period and running optimization scenarios. Various time horizons were used to gain an accurate understanding of the impact. Third, to understand how the initiatives described above improve both revenue and costs, a clear understanding of the profitability of each lens must be considered before JJVC management makes high-level strategic decisions. To make this possible, a Total Delivered Cost (TDC) model was developed and published a for the Contact Lens supply chain.
by Andrew J. Freiheit.
M.B.A.
S.M.
M.B.A. Massachusetts Institute of Technology, Sloan School of Management
S.M. Massachusetts Institute of Technology, Department of Mechanical Engineering

Doctora en Ciencias de la Ingeniería, Mención Química
Productos biotecnológicos, como los biofarmacéuticos entre otros, son productos cuyas tecnologías de producción están en constante desarrollo. Adicionalmente, sus escalas de producción son pequeñas haciendo de las plantas batch las más apropiadas para su producción. En particular, plantas batch multi-producto permiten la producción de una variedad de productos biotecnológicos con varias etapas en común. Una forma de modelar el diseño de plantas batch multi-producto es mediante el enfoque basado en optimización que fue estudiado por primera vez para este tipo de plantas por Robinson y Lonkar, quienes estudiaron el diseño de este tipo de plantas dimensionando los equipos que la conforman. Pese a los múltiples avances en el área, los que incluyen decisiones como la duplicación de unidades, la disposición de tanques de almacenamiento intermedio, programación de la producción y consideración ambientales, entre otras mejores, aún existe una falta de trabajos donde este tipo de enfoques es aplicado en plantas reales. En este trabajo se estudia una reformulación Entera-Mixta Lineal (MILP) del problema Entero-Mixto No-Lineal (MINLP) que resulta al plantear el modelo para el diseño de una planta biotecnológica batch multi-producto. En un primer paso se estudia una reformulación MILP que permite modelar el diseño de una planta utilizando tamaños de equipos en un conjunto continuo y una selección de hosts en un conjunto discreto de opciones. Esta reformulación hace uso de técnicas avanzadas de reformulación, probando ser escalable y confiable para su aplicación en casos reales. En un segundo paso, la reformulación MILP original fue modificada para la inclusión de una selección de equipos, tanto en un conjunto discreto, como en uno continuo, dando un enfoque más realista para poder modelar una planta biotecnológica batch multi-producto; donde unidades como los reactores pueden ser construidos de acuerdo con las necesidades del cliente, sin embargo, unidades como las columnas cromatográficas sólo están disponibles en tamaños discretos dados por el proveedor. Información de procesos reales que formaban parte de una planta batch multi-producto real permitieron la determinación de los parámetros del modelo y una comparación entre las distintas lineas de producción versus la planta real mostraron que este tipo de modelos puede permitir grandes ahorros en los costos de los principales equipos de la planta. Finalmente, como el enfoque estudiado utiliza software de modelación y optimización, el modelo es más amigable para quienes puedan utilizarlo en la práctica. Sin embargo niveles más bajos de implementación podrían mejorar los tiempos de resolución permitiendo la inclusión de formulaciones más complejas, como por ejemplo, la inclusión de costos u objetivos de producción variables.

Volvo Trucks is one of the world’s largest truck manufacturers, with a cab production plant located in Umeå, Sweden. The variation in batch sizes of the different components causes variation in the length of the production plan of subparts, and thus variation in production lead times. This project aims to examine how equally sized batches affect the length of the production plan, and what batch size is optimal to achieve efficient production planning. The examination was conducted with respect to lean principles and a mathematical model was built to simulate the use of different batch sizes. In order to run the simulation, both historical and new data was used. Parameters interesting to examine for the result are the length and variation of the production plan, number of set-ups, production frequency, inventory levels and the annual production. A batch size of 2.0 hours is optimal as the length of the production plan then varies the least in the allowed interval. Equally sized batches exclusively were not found to contribute to a more efficient production planning. Smaller batch sizes in combination with equally sized batches were however shown to decrease the variation in the production plan, and to result in increased stock turnover and decreased inventory levels.
Volvo Lastvagnar är en av världens största lastbilstillverkare, med en fabrik för tillverkning av hytter placerad i Umeå. Variationen i orderstorlekar för olika komponenter orsakar variation i längden av produktionsplanen, vilket i sin tur leder till varierade ledtider. Detta projekt syftar till att undersöka hur utjämnade orderstorlekar påverkar längden av produktionsplanen och vilken orderstorlek som är optimal för en effektiv produktionsplanering. Projektet utfördes med hjälp av leanprinciper och en matematisk modell av produktionen byggdes för att simulera användningen av olika orderstorlekar. För att köra simuleringen användes både historiskt och nytt data. Intressanta parametrar för resultatet är längden och variationen i produktionsplan, antal verktygsbyten, produktionsfrekvens, lagernivåer och årlig produktion. En orderstorlek på 2,0 timmar är optimal eftersom längden på produktionsplanen då har minst variation inom det tillåtna intervallet. Enbart jämna orderstorlekar visade sig inte bidra till en effektivare produktionsplanering, men mindre orderstorlekar i kombination med jämna orderstorlekar visade sig minska variationen i produktionsplan och resulterade i ökad lageromsättning samt minskade lagernivåer.

This work focuses on the application of neural networks in the areas of modelling, identification, control and optimization of biothechnology processes, mainly fed-batch bioreactors. The basic ideas and techniques of artificial neural networks are presented with the notation familiar to control engineers. The applications of a variety of neural network architectures in control and control schemes are first surveyed. Some especific fed-batch bioreactor processes are mentioned to illustrate particular control cases to be examimined in detail and solved. Especifically, a non-linear multivariable bioreactor control problem is used as a case study for model based control techniques. An implementation of direct and inverse process control models based on neural networks that considers biological, thermal and pH effects for this multivariable fed-batch bioreactor is performed and tested. Multilayer perceptrons and radial basis functions neural networks are considered to model this type of non-linear multi-input multi-output (MIMO) dynamic process. The direct models are successfully tested under steady state, dynamic process operation and when a acid disturbance in the process causes a plant/model mismatch. The inverse process model is also successfully tested at the set-point input with a random series of perturbations around the plant operation state. The RBF architecture with goal 3.0 is the best architecture for the direct model of this multivariable process while the best inverse model is based on a MLP 19-11-7-1 trained including past information of the steady states of the process.
On the other hand, optimal control techniques that employ neural networks are studied to optimize the production of invertase in a fed-batch bioreactor. The controlled addition of substrates is used in this bioreactor process to increase productivity when end-product inhibition or catabolite repression are present. Cloned invertase production in Saccharomyces cerevisiae yeast is carried out in fed-batch mode of operation because the enzyme expression is repressed at high glucose concentrations. An optimal glucose feed rate profile is needed to achieve the highest fermentation profit. The controller has to find at each time step an optimal control action that increments the fed-batch bioreactor profitability, even when a disturbance or a set-point change arise. This optimal control action increases the productivity and, within the same optimization process, finds the optimal fermentation ending time. This double optimization is a novelty not met by previous optimization schemes published in the literature. A neuro dynamic programming (NDP) approach coupled with MLP neural networks or fuzzy ARTMAP systems is employed to accomplish these optimization objectives. Fuzzy ARTMAP creates multidimensional category maps by incremental supervised learning. The optimization method utilizes suboptimal control policies as a starting guess. The neural networks are used to build a cost surface in the state space visited by the process. Bellman's iteration is used to improve the cost approximation. The cost surface obtained is implemented into a control system. The controller is tested for different fermentation processes started with different initial fermentation volumes. NDP outperforms other optimization methods employed to find an optimal feeding profile. Besides, it can be used to optimize any fermentation process (starting at different initial conditions) because the future costs (profits) are characterized as a function of system states. The optimal control trajectories found by the controller are similar to the best suboptimal policy for each initial volume. MLP-NDP controllers yield the highest profits, but the manipulated variable trajectories are not smooth. Fuzzy ARTMAP-NDP overcomes this limitation. The best fuzzy ARTMAP-NDP based control system is also tested when an abrupt death of yeast cells occurs. In this case, the controller performance is better than the performance of the fermentation using the best suboptimal policy for the given initial volume.
The integration of control science with neural networks in a unified presentation and identification key areas is a path to follow in future research. Artificial neural networks techniques can be succesfully applied to control fed-batch bioreactors.
Optimización y Control de Procesos de Fermentación Fed-batch a través de Sistemas Neuronales Artificiales
Los procesos de fermentación son ampliamente utilizados en la industria química, farmacéutica y alimentaria. La producción de comida para animales, yogures, quesos, cerveza, colorantes para comida, fertilizantes, medicinas terapéuticas y penicilina, entre otros, son algunos ejemplos de procesos biotecnológicos. En una búsqueda rápida en Internet, cerca de 113000 sitios fueron encontrados, todos ellos relacionados con productos de procesos de fermentación.
El campo de aplicación de la bioingeniería abarca desde los procesos tradicionales de fermentación de vinos hasta la industrialización de no solo la producción de cerveza, queso y leche; sino también de nuevos productos biotecnólogicos como son los antibióticos, enzimas, hormonas, vitaminas, azucares y ácidos orgánicos. Desde tiempos remotos los microorganismos han sido utilizados por el hombre en la producción de alimentos esenciales, como el pan o el queso. El arte de hacer vino ha pasado de padre a hijo desde el antiguo Egipto. Solo hasta finales del siglo XIX, gracias a los estudios del químico y microbiólogo Louis Pasteur, nació la biotecnología como ciencia. La definición actual de biotecnología, según la OCDE, es la aplicación de la ciencia y la tecnología tanto a organismos vivos como a partes, productos y modelos de ellos, con el propósito de modificar tanto materia viva como la no viva con el fin de producir conocimiento, bienes y servicios.
Un proceso de fermentación es un proceso químico en el cual se emplean microorganismos para obtener un producto en particular, aprovechando la selectividad de los microorganismos para producir un determinado compuesto. Los procesos de fermentación son llevados a cabo en un bioreactor. Un bioreactor es un recipiente en el cual microorganismos son cultivados de manera controlada y/o materia prima es convertida o transformada debido a reacciones biológicas.
En una fermentación, es necesario un control adecuado de todas las variables de proceso, debido a que cualquier cambio inesperado en el valor de alguna de ellas puede afectar el desarrollo de los microorganismos y en consecuencia disminuir la productividad del bioreactor. El principal objetivo de controlar un proceso de fermentación es maximizar la producción de microorganismos u otros compuestos metabólicos. Avances recientes en ingeniería genética han aumentado la importancia del adecuado control de los procesos biotecnólogicos. El uso de células de mamíferos o microorganismos en la producción de moléculas complejas requiere el análisis y control de todas las variables de proceso, tales como temperatura, concentración de oxigeno y pH.
El presente trabajo se centra en la aplicación de redes neuronales artificiales en las áreas de modelado, identificación, control y optimización de procesos biotecnológicos, principalmente en bioreactores del tipo fed-batch. Un bioreactor fed-batch se emplea cuando la producción de determinado compuesto de interés, es inhibida debido a la alta concentración de substrato. En un proceso fed-batch, la fermentación empieza con un volumen, concentración de microorganismos y substrato determinados y, a medida que transcurre el proceso de fermentación, el substrato se añade poco a poco, hasta que se consigue llenar el bioreactor.
En este trabajo las ideas y técnicas utilizadas por las redes neuronales artificiales son presentadas con la notación familiar para un ingeniero de control. Diferentes estructuras de redes neuronales artificiales y su posible aplicación a diferentes sistemas de control son resumidas. También son presentados algunos procesos de fermentación fed-batch. Dichos procesos son empleados para ilustrar casos específicos de problemas control. Específicamente, un modelo no lineal y multivariable de un bioreactor es empleado para ilustrar las técnicas de control basadas en el modelo del proceso. Un modelo para la producción de invertasa a través de la levadura Saccharomyces cerevisiae es empleado para ilustrar las técnicas de optimización y control.
Dentro de las técnicas de control basadas en el modelo del proceso, se implementó un modelo directo y uno inverso de la fermentación multivariable antes mencionada. Ambos modelos, basados en redes neuronales artificiales, consideran efectos biológicos, térmicos y de pH. Multilayer perceptron y Radial Basis Function son las redes neuronales empleadas para la construcción de los dos modelos. Para ilustrar la fiabilidad de estos modelos, diferentes pruebas les fueron realizadas. El modelo directo del proceso de fermentación, basado en redes neuronales, fue probado en operación en estado estacionario, en estado dinámico y cuando una perturbación en el ácido causa que el pH del proceso sea diferente. El modelo inverso del proceso de fermentación fue probado haciendo cambios aleatorios en el punto de referencia. Con la arquitectura Radial Basis Function se obtuvo el mejor modelo directo. Para el modelo inverso del proceso de fermentación, se encontró que la mejor arquitectura es la multilayer perceptron 11-7-1, entrenada con informacion de los estados estacionarios del proceso.
Por otro lado, para la optimización de la producción de invertasa es necesario encontrar el perfil de alimentación optimo, de manera que la productividad del bioreactor sea máxima y el tiempo de fermentación sea mínimo. Este doble objetivo de optimización es una novedad y no ha sido antes obtenido por otros esquemas de optimización previamente publicados. El objeto del controlador debe ser hallar a cada instante de tiempo la acción optima de control, es decir, cada vez encontrar cual es el flujo de alimentación adecuado para cumplir el doble objeto de la optimización. En este trabajo se utiliza la programación dinámica neuronal (NDP) con el fin de implementar dicho controlador. Esta técnica emplea redes multilayer perceptron o fuzzy ARTMAP para realizar la optimizacion del proceso. NDP utiliza perfiles de alimentación subóptimas como suposición inicial. A través de esta suposición, una red neuronal es empleada para construir la superficie de costos en el espacio de los estados del proceso. Esta superficie de costos se mejora a través de la iteración de Bellman. Una vez obtenida una buena aproximación a la superficie de costos óptima, esta es implementada en un sistema de control que hace uso también de la ecuación de Bellman. El controlador es probado en diferentes condiciones de operación del proceso de fermentación, específicamente cuando la fermentación comienza con diferentes volúmenes iniciales. Al comparar la metodología empleada se encontró que esta es mejor que otros métodos de optimización utilizados con el mismo fin, debido a que la metodología NDP puede ser usada en diferentes procesos de fermentación sin necesidad de realizar una optimización on-line. Las trayectorias óptimas encontradas por el controlador son similares a la trayectoria seguida por el mejor de los perfiles subóptimos. Con multilayer perceptron- NDP se obtienen los mas altos rendimientos pero la trayectoria de variable manipulada es muy abrupta. Con fuzzy ARTMAP-NDP no se presenta este problema. El controlador que implementa fuzzy ARTMAP-NDP es probado también cuando hay un cambio brusco en la concentración de células. El 50% mueren. En este caso el desempeño del controlador es mejor que el rendimiento de la fermentación cuando la mejor de los perfiles de alimentación subóptimas es utilizado.
Por último se puede decir que la integración de la ingeniería de control con las redes neuronales es un fructífero camino a seguir por futuras líneas de investigación ya que las redes neuronales artificiales pudieron ser empleadas con éxito en el control de bioreactores fed-batch.

Publicaciones

Valencia C., Giralt J., Arenas A., Giralt F., Implementation of a non-linear multivariable (MIMO) process control model of a fed-batch bioreactor with neural networks. Poster Session: Topics in systems and process control, AIChE annual meeting Reno 2001.

Valencia C., Giralt J., Arenas A., Giralt F., Non-linear multivariable (MIMO) process control model of a fed-batch bioreactor with neural networks. Submitted to Chemical Engineering Science. 2002

Valencia C., Lee J.H., Kaisare N.S., Final time and productivity optimization of a fed-batch bioreactor for invertase production. Presented at Control of Pharmaceutical and Biological Processes Session 347, AIChE annual meeting, Indianapolis 2002.

Valencia C., Giralt J., Arenas A., Giralt F., Optimization of invertase production in a fed-batch bioreactor using dynamic programming coupled with fuzzy ARTMAP, to be submitted to Biotechnology and Bioengineering 2002
Optimització i Control dels Processos de Fermentació Fed-batch a través de Sistemes Neuronals Artificials
Els processos de fermentació són amplament utilitzats en l'indústria química, farmacèutica i alimentaria. La producció de menjar per a animals, iogurts, formatge, cervesa, colorants per aliments, fertilitzants, medicines terapèutiques i penicil·lina entre altres, són alguns exemples de processos biotecnològics. En una recerca ràpida a Internet, cerca de 113000 llocs van ser trobats, tots ells relacionats amb productes de processos de fermentació.
El camp d'aplicació de la bioenginyeria avarca des d'els tradicionals processos de fermentació de vi fins a l'industrialització de no solament la producció de cervesa, formatge i llet; sinó també de noves productes biotecnològics com són els antibiòtics, enzims, hormones, vitamines, sucres i àcids orgànics. Des de temps remots els microorganismes van ser utilitzats per l'home en la producció d'aliments bàsics, com el pa o el formatge. L'art de fer vi a passat de pares a fills des d'l antic Egipte. Només fins a finals del segle XIX, gràcies als estudis del químic i microbiòleg Louis Pasteur, va nàixer la biotecnologia com a ciència. La definició moderna de biotecnologia, segons la OCDE, és l'aplicació de la ciència i la tecnologia tant als organismes vius com a les seves parts, productes i models d'ells, amb el propòsit de modificar tant la matèria viva com la no viva amb l'objecte de produir coneixements, bens i serveis.
Un procés de fermentació és un procés químic que fa servir microorganismes per a obtenir un producte en particular, aprofitant la selectivitat dels microorganismes per a produir un determinat compost. Els processos de fermentació es realitzen en un bioreactor. Un bioreactor és un vaixell on els microorganismes són cultivats de forma controlada i/o matèria primera és convertida o transformada per reacciones biològiques.
En una fermentació, es necessita un adequat control de totes les variables de procés, per tal que qualsevol canvi inesperat en el valor d'alguna de elles pot afectar el desenvolupament de els microorganismes i en conseqüència disminuir la productivitat del bioreactor. El principal objectiu de controlar un procés de fermentació, és maximitzar la producció de microorganismes o altres compostos metabòlics. Recents avanços en enginyeria genètica han augmentat la importància de l'adequat control dels processos biotecnològics. L'ús de cèl·lules de mamífers o microorganismes en la producció de molècules complexes necessita de l'anàlisi i el control de totes las variables de procés, tal com temperatura, concentració d'oxigen i pH.

El present treball es centra en l'aplicació de xarxes neuronals artificials en les àrees de modelat, identificació, control i optimització de processos biotecnològics, principalment en bioreactors de tipus fed-batch. Un bioreactor fed-batch es fa servir quan la producció d'un determinat compost d'interès, és inhibida per l'alta concentració de substrat. En un procés fed-batch, la fermentació comença amb un volum, concentració de microorganismes i substrat determinats i a mida que transcorre el procés de fermentació, el substracte s'agrega poc a poc, fins que el bioreactor és ple.
En aquell treball les idees i tècniques utilitzades per las xarxes neuronals artificials són presentades amb la notació familiar per a un enginyer de control. Diferents estructures de xarxes neuronals artificials i la seva possible aplicació a diferents sistemes de control van ser resumides. També s'han presentat alguns processos de fermentació fed-batch. Aquests processos es fan servir per il·lustrar casos específics de problemes de control. Específicament, un model no lineal i multivariable d'un bioreactor es fa servir per il·lustrar les tècniques de control basades en el model del procés. Un model per a la producció de invertasa a través del llevat Saccharomyces cerevisiae es fa servir per il·lustrar les tècniques d'optimització i control.
Dins les tècniques de control basades en el model del procés, es va a implementar un model directe i un invers de la fermentació multivariable. Els dos models, basats en xarxes neuronals artificials, consideren efectes biològics, tèrmics i de pH. Multilayer perceptron i Radial Basis Function són las xarxes neuronals que es van a fer servir per a la construcció d'ambdós models. Per il·lustrar la fiabilitat d'aquests models, diferent proves van ser realitzades. El model directe del processo de fermentació, basat en xarxes neuronals, va ser provat quan el procés opera en estat estacionari, en estat dinàmic i quan una perturbació en l'àcid causa que el pH del procés sigui un altre. El model invers del procés de fermentació també va ser provat fent canvis aleatoris del punt de consigna. L'arquitectura Radial Basis Function va a ser el millor model directe que es va a trobar. Pel model invers del procés de fermentació, es va a trobar que la millor arquitectura es la Multilayer Perceptron 11-7-1, que va ser entrenada amb informació dels estats estacionaris del procés.
Per una altre banda, per a l'optimització de la producció de invertasa es necessita trobar el perfil d'alimentació òptim, de manera que la productivitat del bioreactor sigui màxima i el temps de fermentació sigui mínim. Aquest doble objectiu de optimització constitueix una novetat i no ha estat obtingut per altres mètodes d'optimització prèviament publicats. L'objectiu del controlador és trobar a cada instant de temps l'acció òptima de control, és dir, cada vegada trobar quin és el flux d'alimentació correcte per complir el doble objecte de l'optimització. Aquest treball va fer servir la programació dinàmic neuronal (NDP) amb l'objectiu de implementar aquell controlador. Aquesta tècnica fa servir xarxes multilayer perceptron o fuzzy ARTMAP. Aquest mètode d'optimització utilitza perfils d'alimentació subòptimes com a suposició inicial. A través d'ella, una xarxa neuronal es utilitzada per construir la superfície de costos en l'espai dels estats del procés. Aquesta superfície de costos es millora a través de la iteració de Bellman. Una vegada obtinguda una bona aproximació a la superfície de costos òptima, aquesta es implementada en un sistema de control que fa ús de l'equació de Bellman. Aquest controlador és provat dins diferentes condiciones d'operació del procés de fermentació, específicament quan la fermentació comença amb diferents volums inicials. S'ha trobat que la metodologia emprada és millor que altres mètodes de optimització ja que es pot utilitzar en altres processos de fermentació sense la necessitat de fer una optimització on-line. Les trajectòries òptimes trobades pel controlador son similars a la trajectòria seguida pel millor dels perfils subòptims. Amb Multilayer Perceptron- NDP s'han obtingut els millores rendiments, però la trajectòria de variable manipulada és força abrupta. Amb Fuzzy ARTMAP-NDP no es presenta aquest problema. El controlador que implementa fuzzy ARTMAP-NDP és provat també quan es presenta un canvi brusc en la concentració de cèl·lules. El 50% moren. En aquell cas el desenvolupament del controlador és millor que el rendiment de la fermentació quan el millor dels perfils d'alimentació subòptimes es fan servir.

Per últim es pot dir que la integració de l'enginyeria de control amb les xarxes neuronals és un camí a seguir per futures línees d'investigació. Las xarxes neuronals artificials poden, amb èxit, fer-se servir en el control de bioreactors fed-batch.

Modeling and inferential control studies are carried out on a reactive batch distillation system for the esterification reaction of ethanol with acetic acid to produce ethyl acetate. A dynamic model is developed based on a previous study done on a batch distillation column. The column is modified for a reactive system where Artificial Neural Network Estimator is used instead of Extended Kalman Filter for the estimation of compositions of polar compounds for control purposes. The results of the developed dynamic model of the column is verified theoretically with the results of a similar study. Also, in order to check the model experimentally, a lab scale column (40 cm height, 5 cm inner diameter with 8 trays) is used and it is found that experimental data is not in good agreement with the models&rsquo
. Therefore, the model developed is improved by using different rate expressions and thermodynamic models (fi-fi, combination of equations of state (EOS) and excess Gibbs free energy (EOS-Gex), gama-fi) with different equations of states (Peng Robinson (PR) / Peng Robinson - Stryjek-Vera (PRSV)), mixing rules (van der Waals / Huron Vidal (HV) / Huron Vidal Original (HVO) / Orbey Sandler Modification of HVO (HVOS)) and activity coefficient models (NRTL / Wilson / UNIQUAC). The gama-fi method with PR-EOS together with van der Waals mixing rule and NRTL activity coefficient model is selected as the best relationships which fits the experimental data. The thermodynamic models
EOS, mixing rules and activity coefficient models, all are found to have very crucial roles in modeling studies. A nonlinear optimization problem is also carried out to find the optimal operation of the distillation column for an optimal reflux ratio profile where the maximization of the capacity factor is selected as the objective function. In control studies, to operate the distillation system with the optimal reflux ratio profile, a control system is designed with an Artificial Neural Network (ANN) Estimator which is used to predict the product composition values of the system from temperature measurements. The network used is an Elman network with two hidden layers. The performance of the designed network is tested first in open-loop and then in closed-loop in a feedback inferential control algorithm. It is found that, the control of the product compositions with the help of an ANN estimator with error refinement can be done considering optimal reflux ratio profile.

The research focuses on investigation and optimization (using Box Behnken design) of the key parameters of CO2 gas concentration, light intensity, temperature, feedwater nutrient concentration, and wastewater origin (municipal primary and secondary, and petroleum industry) on photobioreactor algal growth parameters. A mathematical model is provided for predicting algal growth. Finally, the influence of light wavelength on algal growth is investigated. The outcomes can be used to inform design and operation of large-scale algal cultivation systems.

La industria química ha experimentado en las últimas décadas un aumento en la competencia por la cual las empresas se ven obligadas a adaptarse a un mercado cambiante y cada vez más exigente. Aunque la globalización ha abierto nuevos mercados, ha incrementado también el número de competidores, de tal manera que sólo las empresas que usen las plantas más integradas y eficientes podrán mantenerse en el negocio. En este contexto global, el principal propósito de esta tesis es desarrollar métodos que exploten la flexibilidad de los procesos, con el objetivo de aumentar la eficiencia de las plantas y asegurar los requerimientos de seguridad y calidad de los productos. Esta tesis contribuye a la optimización y a la gestión de la producción desde pequeñas plantas que usen procesos discontinuos hasta grandes plantas de procesado continuo.
En primer lugar, esta tesis trata la gestión de los procesos continuos en los que suelen fabricar productos muy similares a gran escala. La gran ventaja de los procesos continuos es que pueden conseguir mayor consistencia en la calidad de los productos y que pueden aprovechar las economías de escala que reducen los costes y residuos. Sin embargo, la industria química para mantenerse competitiva necesita adaptar continuamente sus procesos a las condiciones del mercado y de operación. El sistema de control supervisor presentado en esta parte de la tesis disminuye el tiempo de reacción frente a incidentes en los procesos continuos y re-optimiza la producción en tiempo real, si existe posibilidad de mejora.
A continuación, esta tesis trata la gestión de los procesos semicontinuos que permiten una operación más flexible y personalizada. Los procesos semicontinuos operan con puestas en marcha y paradas periódicas para acomodar las frecuentes transiciones entre diferentes productos. Esta tesis presenta un nuevo concepto de fabricación flexible que permite programar perfiles variables de velocidad de producción dentro de cada campaña de producción.
La mayor parte del trabajo de investigación de esta tesis se dedica a la planificación de la producción en los procesos discontinuos por lotes, utilizados principalmente en la producción de productos químicos con alto valor añadido. Estos procesos ofrecen varias ventajas respecto a los procesos continuos y semicontinuos debido a la mayor flexibilidad para acomodar diversos productos, diferentes capacidades de producción, y la posibilidad de realizar operaciones completamente diferentes en los mismos equipos. Sin embargo, la obtención del plan de producción óptimo usando se complica al aumentar la complejidad de la planta y/o el número de lotes a planificar. La simplificación de considerar tiempos de transferencia despreciables es generalmente aceptada en la literatura para evitar la complejidad del manejo de las operaciones de transferencia. En cambio, esta tesis pretende resaltar el papel crítico que juegan las operaciones de transferencia en la sincronización de tareas, y en la consiguiente determinación de planes de producción factibles.
Siguiendo con los procesos por lotes, esta tesis demuestra que el uso del concepto de recetas flexibles mejora la operación de los procesos en ambientes de producción con mucha incertidumbre. La flexibilidad de las receta se considera como una oportunidad adicional, tanto para la planificación de la producción reactiva como preactiva, reduciendo el riesgo de llegar a resultados económicamente desfavorables.
Finalmente, esta tesis presenta las plantas discontinuas sin tuberías como una alternativas a las plantas por lotes clásicas. En la búsqueda de formas más competitivas y efectivas de producción, la flexibilidad para producir un elevado número de productos en plantas por lotes es limitada debido a la necesidad de equipos fijos conectados por tuberías y frecuentes tareas de limpieza. Las plantas sin tuberías presentan una mayor flexibilidad ya que el material se transfiere entre estaciones de procesamiento usando equipos que se mueven dentro de la planta. El trabajo presentado en esta parte de la tesis contribuye a la mejora en la gestión de este tipo de plantas proponiendo una formulación más eficiente a las encontradas en la literatura que resuelve el problema de la planificación de la producción.
En resumen, esta tesis desarrolla nuevas estrategias de modelado y métodos de resolución encaminados al soporte de la toma de decisiones que explotan la flexibilidad intrínseca de los procesos químicos. Las principales ventajas de cada una de las contribuciones de esta tesis se demuestran mediante su aplicación a diferentes casos de estudio.
The chemical industry has become increasingly competitive over the past decades. Companies are required to adapt to changing market conditions and meet stricter product specifications. While globalization has opened new markets for the chemical industry, it has also increased the competitor pool, giving an advantage to companies with more efficient and highly integrated plants.
In this context, the main aim of this thesis is to demonstrate new concepts and computational methods that exploit process flexibility to enhance plant profitability under transient operating conditions. These methods ensure that safety and product quality requirements are consistently met. This thesis makes contributions to the optimization and management of production in plants ranging from small batch plants to large capacity continuous processes.
First, this thesis addresses the management of continuous processes, in which similar products are mass produced. Continuous processes can achieve the highest consistency and product quality by taking advantage of economies of scale and reduced manufacturing costs and waste. However, in order to remain competitive in the market, plants are required to dynamically adapt their processes to fit the continuously changing market and operating conditions. The supervisory control system presented in this part of the thesis decreases the system reaction time to incidences and re-optimizes the production in real time if the opportunity for improved performance exists.
Next, this thesis addresses the management of semicontinuous processes, which allow more customized and flexible operation. Semicontinuous processes run with periodic start-ups and shutdowns to accommodate frequent product transitions. This thesis proposes an optimization model that creates improved production schedules by introducing a new concept of flexible manufacturing that allows production rate profiles to be programmed within each operation campaign.
The major part of the research work of this thesis deals with the operational management of batch processes, which are mainly used for the production of high value-added chemicals. Batch processing offers the advantage of increased flexibility in product variety, production volume, and the assortment of operations that can be processed by a particular piece of equipment. However, the trade-off is that production scheduling is significantly complicated by the large number of batches involved with different production paths. In order to avoid the complexity of managing transfer operations, the assumption of negligible transfer times is generally accepted in batch scheduling. Conversely, this thesis highlights the critical role that transfer operations play in the synchronization of tasks and in determining the feasibility of production schedules.
Continuing to focus on batch plant operation, this thesis demonstrates that the use of the concept of flexible recipes enhances the operation batch plants within an uncertain environment. Recipe flexibility is considered as an additional opportunity for reactive scheduling as well as a proactive way to reduce the risk of meeting unfavorable scenarios.
Finally, this thesis examines pipeless plants as an alternative to batch plants. In the search for more competitive and effective ways of production, flexibility of batch plants for producing a large number of products is limited due to the need for equipment, piping and frequent cleaning tasks. Pipeless plants have enhanced flexibility over batch plants, because the material is moved along its production path through moveable vessels. This part of the thesis contributes to the optimization of the management of pipeless plants by proposing an alternative formulation for solving short-term scheduling problems.
In summary, this thesis provides novel modeling approaches and solution methods aimed at supporting the decision-making process in plant production scheduling which exploit the existing flexibility in chemical processes. The main advantages of each contribution are highlighted through case studies.

Monoclonal antibodies (MAbs) have an expanding market for use in diagnostic and therapeutic applications. Industrial production of these biopharmaceuticals is usually achieved based on fed-batch cultures of mammalian cells in bioreactors (Chinese hamster ovary (CHO) and Hybridoma cells), which can express different kinds of recombinant proteins. In order to reach high cell densities in these bioreactors, it is necessary to carry out an optimization of their production processes. Hence, macroscopic model equations must be developed to describe cell growth, nutrient consumption and product generation. These models will be very useful for designing the bioprocess, for developing robust controllers and for optimizing its productivity.

This thesis presents a new kinetic model of hybridoma cell metabolism in fed batch culture and typical illustration of a systematic methodology for mathematical modelling, parameter estimation and model-based optimization and state estimation of bioprocesses.

In the first part, a macroscopic model that takes into account phenomena of overflow metabolism within glycolysis and glutaminolysis is proposed to simulate hybridoma HB-58 cell cultures. The model of central carbon metabolism is reduced to a set of macroscopic reactions. The macroscopic model describes three metabolism states: respiratory metabolism, overflow metabolism and critical metabolism. The model parameters and confidence intervals are obtained via a nonlinear least squares identification. It is validated with experimental data of fed-batch hybridoma cultures and successfully predicts the dynamics of cell growth and death, substrate consumption (glutamine and glucose) and metabolites production (lactate and ammonia). Based on a sensitivity analysis of the model outputs with respect to the parameters, a model reduction is proposed.

In the next step, the effort is directed to the maximization of biomass productivity in fed-batch cultures of hybridoma cells based on the overflow metabolism model. Optimal feeding rate, on the one hand, for a single feed stream containing both glucose and glutamine and, on the other hand, for two separate feed streams of glucose and glutamine are determined using a Nelder-Mead simplex optimization algorithm. Two different objective functions (performance criteria) are considered for optimization; the first criterion to be maximized is the biomass productivity obtained at the end of the fed-batch culture, the second criterion to be minimized is the difference between global substrate consumption and the maximum respiratory capacity.

The optimal multi exponential feed rate trajectory improves the biomass productivity by 10% as compared to the optimal single exponential feed rate. Moreover, this result is validated by the one obtained with the analytical approach in which glucose and glutamine are fed to the culture so as to control the hybridoma cells at the critical metabolism state, which allows maximizing the biomass productivity. The robustness analysis of optimal feeding profiles obtained with different optimization strategies is considered, first, with respect to parameter uncertainties and, finally, with respect to model structure errors.

Finally, the overflow metabolism model is used to develop an extended Kalman filter for online estimation of glucose and glutamine in hybridoma cell fed-batch cultures based on the considered available measurements (biomasses (on-line), lactate and ammonia (on-line or off-line)). The observability conditions are examined, and the performances are analysed with simulations of hybridoma cell fed-batch cultures. Glutamine estimation sensitivity is enforced by minimizing a cost function combining a usual least-squares criterion with a state estimation sensitivity criterion.

Doctorat en Sciences de l'ingénieur
info:eu-repo/semantics/nonPublished

In this study, the effects of pH on therapeutically important protein, recombinant human erythropoietin (rhuEPO), production by Pichia pastoris was investigated at pH=4.0, 4.5, 5.0, 5.5 and 6.0. rHuEPO production was started by methanol induction in fed-batch mode. The highest cell concentration was obtained at pH=4.5 as 81.4 g L-1. The co-substrate substrate sorbitol, which was added batch-wise, was consumed at t=15 h of the operations at pH=4.0, 4.5 and 5.0. However as the pH increases above pH=5.0 the sorbitol consumption rate decreases. The highest rHuEPO concentration was achieved at pH=4.5 as 0.158 g L-1 which was 1.43-, 1.24-, 1.95- and 1.23-fold higher than those obtained at pH=4.0, 5.0, 5.5, and 6.0, respectively. Also at pH=4.5 overall cell yield on substrate was 0.51 g g-1 and overall rHuEPO yield on substrate was 1.45 mg g-1. rHuEPO concentration was decreased in the last 3-6 hour of the operation due to proteolysis. Therefore extracellular protease concentrations in the medium were determined. As expected, since the investigated pH range was acidic, the amount of acidic proteases was found to be higher than neutral and basic proteases. Furthermore the total protease concentration increased linearly in the fermentation broth, having close values at different pH values. Thus, pH did not have a significant effect on extracellular protease activity. Alcohol oxidase (AOX) activities showed similar behavior at different pH. The highest specific AOX activity was attained at pH=4.5, at which the highest rHuEPO concentration was achieved, as 110.1 U g-1 CDW. Keywords:

The increasing trend of world-wide energy consumption emphasizes the importance of ongoing optimization of new and existing technologies. In this dissertation, two energy–intensive systems are simulated and optimized. Advanced estimation, optimization, and control techniques such as a moving horizon estimator and a model predictive controller are developed to enhance the profitability, product quality, and reliability of the systems. An enabling development is presented for the solution of complex dynamic optimization problems. The strategy involves an initialization approach to large–scale system models that both enhance the computational performance as well as the ability of the solver to converge to an optimal solution. One particular application of this approach is the modeling and optimization of a batch distillation column. For estimation of unknown parameters, an L1-norm method is utilized that is less sensitive to outliers than a squared error objective. The results obtained from the simple model match the experimental data and model prediction for a more rigorous model. A nonlinear statistical analysis and a sensitivity analysis are also implemented to verify the reliability of the estimated parameters. The reduced–order model developed for the batch distillation column is computationally fast and reasonably accurate and is applicable for real time control and online optimization purposes. Similar to estimation, an L1-norm objective function is applied for optimization of the column operation. Application of an L1-norm permits explicit prioritization of the multi–objective problems and adds only linear terms to the problem. Dynamic optimization of the column results in a 14% increase in the methanol product obtained from the column with 99% purity. In a second application of the methodology, the results obtained from optimization of the hybrid system of a cryogenic carbon capture (CCC) and power generation units are presented. Cryogenic carbon capture is a novel technology for CO2 removal from power generation units and has superior features such as low energy consumption, large–scale energy storage, and fast response to fluctuations in electricity demand. Grid–level energy storage of the CCC process enables 100% utilization of renewable power sources while 99% of the CO2 produced from fossil–fueled power plants is captured. In addition, energy demand of the CCC process is effectively managed by deploying the energy storage capability of this process. By exploiting time–of–day pricing, the profit obtained from dynamic optimization of this hybrid energy system offsets a significant fraction of the cost of construction of the cryogenic carbon capture plant.

This research is concerned with the development of efficient solutions to various problems that arise in the flow-shop environments which utilize lot-streaming. Lot streaming is a commonly used process of splitting production lots into sublots and, then, of scheduling the sublots in an overlapping fashion on the machines, so as to expedite the progress of orders in production and to improve the overall performance of the production system. The different lot-streaming problems that arise in various flow-shop environments have been divided into two categories, single-lot problems and multiple-lot problems. Further classification of the multiple-lot problems into the lot streaming sequencing problem (LSSP) and the flow-shop lot-streaming (FSLS) problem is made in this work. This classification is motivated by the occurrence of these problems in the industry. Several variants of these problems are addressed in this research. In agreement with numerous practical applications, we assume sublots of equal sizes. It turns out that this restriction paves the way to the relaxation of several typical limitations of current lot-streaming models, such as assumption of negligible transfer and setup times or consideration of only the makespan criterion. For the single-lot problem, a goal programming (GP) approach is utilized to solve the problem for a unified cost objective function comprising of the makespan, the mean flow time, the average work-in-process (WIP), and the setup and handling related costs. A very fast optimal solution algorithm is proposed for finding the optimal number of sublots (and, consequently, the sublot size) for this unified cost objective function in a general m-machine flow shop. For the more complicated multiple-lot problem, a near-optimal heuristic for the solution of the LSSP is developed. This proposed heuristic procedure, referred to as the Bottleneck Minimal Idleness (BMI) heuristic, identifies and employs certain properties of the problem that are irregular in traditional flow-shop problems, particularly the fact that the sublot sizes eminating from the same lot type and their processing times (on the same machines) are identical. The BMI heuristic attempts to maximize the time buffer prior to the bottleneck machine, thereby minimizing potential bottleneck idleness, while also looking-ahead to sequence the lots with large remaining process time earlier in the schedule. A detailed experimental study is performed to show that the BMI heuristic outperforms the Fast Insertion Heuristic (the best known heuristic for flow-shop scheduling), when modified for Lot Streaming (FIHLS) and applied to the problem on hand. For the FSLS problem, several algorithms are developed. For the two-machine FSLS problem with an identical sublot-size for all the lots, an optimal pseudo-polynomial solution algorithm is proposed. For all practical purposes (i.e., even for very large lot sizes), this algorithm is very fast. For the case in which the sublot-sizes are lot-based, optimal and heuristic procedures are developed. The heuristic procedure is developed to reduce the complexity of the optimal solution algorithm. It consists of a construction phase and an improvement phase. In the construction phase, it attempts to find a near-optimal sequence for the lots and then, in the improvement phase, given the sequence, it attempts to optimize the lot-based sublot-sizes of each of the lots. Extensions of the solution procedures are proposed for the general m-machine FSLS problem. A comprehensive simulation study of a flow shop system under lot streaming is conducted to support the validity of the results and to demonstrate the effectiveness of the heuristic procedures. This study clearly indicates that, even in dynamic practical situations, the BMI rule, which is based on the proposed BMI heuristic, outperforms existing WIP rules, commonly used in industry, in scheduling a flow-shop that utilizes lot streaming. With respect to the primary performance measure - cycle time (or MFT) - the BMI rule demonstrates a clear improvement over other WIP rules. It is further shown that it also outperforms other WIP rules with respect to the output variability measure, another important measure in flow-shop systems. The effects of several other factors, namely system randomness, system loading, and bottleneck-related (location and number), in a flow-shop under lot streaming, are also reported.
Ph. D.

The synthesis of a number of alkyl esters such as methyl lactate, methyl decanoate, and ethyl benzoate via esterification in a reactive distillation is quite challenging. It is due to the complexity in the thermodynamic behaviour of the chemical species in the reaction mixture in addition to the difficulty of keeping the reactants together in the reaction section. One of the reactants (in these esterification reactions) having the lowest boiling point can separate from the other reactant as the distillation continues. This can result in a significant drop in the reaction conversion in a conventional reactive distillation whether it is a batch or a continuous column. To overcome this challenge, new different types of batch reactive distillation column configurations: (1) integrated conventional (2) semi-batch (3) integrated semi-batch (4) integrated dividing-wall batch distillation columns have been proposed here. Four esterification reaction schemes such as (a) esterification of lactic acid (b) esterification of decanoic acid (c) esterification of benzoic acid (d) esterification of acetic acid are investigated here. A detailed dynamic model based on mass, energy balances, chemical reaction, and rigorous thermodynamic (chemical and physical) properties is considered and incorporated in the optimisation framework within gPROMS (general PROcess Modelling System) software. It is found that for the methyl lactate system, the i-SBD operation outperforms the classical batch operations (CBD or SBD columns) to satisfy the product constraints. While, for the methyl decanoate system, the i-DWCBD operation outperforms all CBD, DWBD and sr-DWBD configurations by achieving the higher reaction conversion and the maximum product purity. For the ethyl benzoate system, the performance of i-CBD column is superior to the CBD process in terms of product quality, and conversion rate of acid. The CBD process is found to be a more attractive in terms of operating time saving, and annual profit improvement compared to the IBD, and MVD processes for the benzyl acetate system.
The Higher Committee for Education Development in Iraq (HCED)

Aquest treball de tesi s'adreça a la indústria de procés químic d'operació discontínua tot i que també pot resultar profitós per altres situacions de fabricació per lots. Aquests tipus d'indústries són molt variats i s'inclouen en sectors com el de la química fina, el farmacèutic, l'agroalimentari i, en definitiva, l'elaboració d'especialitats. La seva importància econòmica rau tant en el seu significatiu volum de producció com en l'alt valor afegit que sovint tenen els productes.
L'objecte de l'estudi és la gestió dels recursos de fabricació i l'optimització de l'operació de la planta per tal d'ajustar la producció resultant als requeriments de la demanda. La natura discontínua i dinàmica dels processos que hi tenen lloc reclama la presa de decisions en termes d'assignació de tasques a equips (reactors, filtres i altres aparells), de seqüenciació de productes i temporització de les operacions. Es tracta, doncs, d'un problema de programació d'activitats.
El primer pas vers la solució d'aquest problema és la modelització del sistema estudiat. Aquest treball proposa un model d'operació detallat que inclou, a més de la compartició per part dels diferents productes de recursos generals (temps, equips, serveis...), aspectes més específics de la indústria química. Són les necessitats de transferència i emmagatzamatge de productes, tant intermedis com acabats, de natura molt probablement fluïda. I són també, les tasques de preparació i neteja que, depenent de la seqüència de productes, cal aplicar als equips emprats.
La modelització de l'emmagatzemament també considera la possible inestabilitat de certs intermedis i la conseqüent limitació de l'espera abans de ser processats. De fet, la modelització incorpora un nou esquema conceptual que permet la segmentació de les receptes dels diferents productes en sèries de tasques lligades als intermedis estables. Llavors, la descripció d'un programa d'operacions complex mitjançant una sèrie d'ordres de fabricació d'aquests intermedis proporciona una notable simplificació del problema que redueix el número de restriccions a considerar i facilita el procediment d'optimització.
L'optimització mitjançant mètodes rigorosos ja establerts només resulta indicada per problemes molt simples en comparació amb el que es planteja en aquest treball. És per això que s'ha desenvolupat una metodologia heurística general que permet abordar aquest problema patró sota les diferents circumstàncies de cada cas particular. En una primera part es procedeix a la generació d'alternatives de producció considerant els casos possibles i eliminant els ineficients sota una sèrie de criteris preestablerts. Aquestes alternatives s'expressen com seqüències de fabricació d'intermedis estables i posteriorment poden ser millorades sota diferents objectius bé manualment o bé emprant mètodes sistemàtics.
La validació de la metodologia proposada és possible gràcies a la comparació amb casos de la literatura. En proposar exemples de major complexitat, la utilització de mètodes estocàstics es demostra molt eficient per resoldre problemes per als quals és difícil trobar regles de solució. Llavors, la validació d'aquest casos és possible estadísticament.
En conclusió, d'aquest treball en resulta una metodologia general i flexible que s'adapta a la majoria de característiques d'un tipus de problemes que, de moment i en última instància, han de ser resolts de manera particular. La seva complexitat continua reclamant la nostra intuïció.
This thesis concerns batch chemical process industries, although it may be of interest in other batch processing areas. This kind of industry is diverse and includes a variety of sectors such as fine chemicals, pharmaceuticals, food and agriculture industry and speciality manufacturing. The economical significance of this industrial activity resides on both, its large production volume and its high added value.
The thesis is aimed at studying the management of limited resources and the optimization of plant operation in order to have the production meet the demand requirements. The inherent discontinuous and dynamic nature of this kind of processes needs decision-making in terms of unit-to-task assignment (reactors, filters and other apparatuses), in terms of product sequencing and in terms of operation timing.
The first step towards the solution of this problem is system modelling. This work proposes a detailed operation model that includes the general sharing of resources (time, equipment, utilities, etc.), as well as more specific aspects related to the chemical industry. These aspects are the special needs for material transfer and storage, both final and intermediate, which are likely to be of fluid nature. Additionally, these aspects also include the special set-up and cleaning tasks that the equipment units require depending on the sequence of tasks performed.
Storage modelling also considers the different stability of the intermediate substances and the resulting limitation of the waiting time before processing. Thus, the modelling incorporates a novel conceptual framework allowing the segmentation of the recipes into different series of tasks defined by the stable intermediate products. Hence, the description of a complex operation schedule by a sequence of production orders related to these intermediate products allows a significant problem simplification, which results in the reduction of the number of constraints to be considered and enhances the performance of the search.
Optimization via established rigorous methods is only suitable for problems simpler than those addressed in this thesis. This is the reason for the development of a general heuristic methodology. The first part generates production alternative paths through an enumerative procedure that eliminates inefficient options according to given pre-established criteria. These alternative paths are characterized as production sequences of storable intermediate products that are next improved regarding different objective functions either manually or automatically.
The validation of the proposed methodology is achieved through a comparative study using cases reported in the literature. For those more complex cases, the use of stochastic methods has demonstrated to be very effective when scheduling rules are not available. For these cases, the assessment and validation is achieved through statistical analysis.
In conclusion, this thesis results in a general and flexible methodology that accommodates the most of the features of a kind of scheduling and planning problems that finally require particular solution strategies. The more complex the problem, the more the human intuition is to some extend still required.

This research focuses on developing a comprehensive framework for designing and modeling experiments in the presence of multiple sources of competing prior knowledge. In particular, methodology is proposed for process optimization in high-cost, low-resource experimental settings where the underlying response function can be highly non-linear. In the first part of this research, an initial experimental design criteria is proposed for optimization problems by combining multiple, potentially competing, sources of prior information--engineering models, expert opinion, and data from past experimentation on similar, non-identical systems. New methodology is provided for incorporating and combining conjectured models and data into both the initial modeling and design stages. The second part of this research focuses on the development of a batch sequential design procedure for optimizing high-cost, low-resource experiments with complicated response surfaces. The success in the proposed approach lies in melding a flexible, sequential design algorithm with a powerful local modeling approach. Batch experiments are designed sequentially to adapt to balance space-filling properties and the search for the optimal operating condition. Local model calibration and averaging techniques are introduced to easily allow incorporation of statistical models and engineering knowledge, even if such knowledge pertains to only subregions of the complete design space. The overall process iterates between adapting designs, adapting models, and updating engineering knowledge over time. Applications to nanomanufacturing are provided throughout.

Ce travail s’intéresse essentiellement au développement et à l’application d’une méthodologie pour l’optimisation dynamique multicritère d’un procédé discontinu alimenté de production bactérienne d’arômes laitiers. L’étude réalisée comprend principalement quatre parties ; la première décrit l'utilisation des plans d’expériences multicritère, couvrant un large domaine expérimental (pH, température et oxygénation), afin de collecter le maximum d'information sur le comportement physiologique de Lactococcus lactis. La deuxième concerne l'analyse cinétique détaillée de l’effet des combinaisons des facteurs opératoires sur la croissance et l'orientation du métabolisme. La troisième, traite de la construction d'un modèle cinétique en mode discontinu et discontinu alimenté. Dans la dernière, l’outil d’optimisation dynamique multicritère a été validé et appliqué au problème de production d’arômes étudié. Les résultats obtenus pour l’optimisation multicritère sont satisfaisants
The aim of the present work is to develop and apply an innovative approach for multicriteria and dynamic optimization of fed-batch fermentation for the production of lactic aroma. This study is mainly divided in four parts: the first one describes the use of multicriteria experimental design, covering an extended operating conditions domain (pH, temperature and oxygenation), in order to collect large experimental information in batch culture of Lactococcus lactis. The second part involves detailed kinetic analysis of operating factor effect combinations on growth and metabolism orientation. The third part deals with the development of a model to predict the kinetic behaviour of this strain in batch and fed-batch cultures. In the last part, the tool for multicriteria and dynamic optimization was validated and applied to the studied problem of lactic aroma production. The results obtained for multicriteria and dynamic optimization are satisfactory

Simultaneous Localization and Mapping (SLAM) aims to estimate the positions and orientations of the mobile robot and to construct the model of the environment. SLAM can help the robot to plan and execute a collision-free trajectory from the current configuration to the target configuration, so it is essential and critical for the mobile robot’s autonomous navigation and effective task execution. SLAM plays a quite important role in a wide range of application fields, from indoor to outdoor, from industry to military, from terrain, submarine to outer space, etc. In indoor dynamic scenarios where there are moving objects, robust SLAM is also important for the mobile robot to co-exist with humans safely and to improve the capability in robot’s estimation for its own state of pose and the surrounding world model. The research goal of this dissertation is to design, implement and validate Graph-SLAM algorithms for mobile robots in indoor office-like dynamic scenarios. Graph-SLAM belongs to the category that addresses the issues of localization and mapping in a hierarchical way, where a topological graph is constructed to represent the robot poses, the local relative motion constraints between them are estimated as the edges of the graph and the global consistent registration is performed to estimate the trajectory of the robot. Graph-SLAM can lead to much accurate results approaching the ground truth. The overview of the designed and implemented Graph-SLAM algorithm includes the following three parts: scan matching to estimate the local relative roto-translation, batch optimization to estimate the global mobile robot’s trajectory, and the global line-feature-based mapping to construct the global line-feature model of the environment. The details of each chapter are briefly shown as follows: On the local level, moving-object-detection based scan matching is accomplished: first, conditioned-Hough-Transform-based segmentation is performed to extract and group the small-scale line-feature-candidate samples; second, occupancy-analysis-based moving-object detection is executed to detect and discard the segments corresponding to the moving objects; third, linear-regression-based line-feature matching is applied to merge the similar small-scale line features into larger-scale line features, and also to match the larger-scale line features in order to estimate the roto-translation values. The experiments will prove the effectiveness of the algorithm to estimate the relative roto-translation value even faced with the disturbances of the moving objects in the dynamic scenario. On the global level, the motion constraints computed from scan matching between the immediate consecutive, the close-by-but-not-adjacent robot poses are used to construct the topological graph, and the least-square cost function associated with the graph is optimized by a linear solution. The experimental tests dealing with the publicly available dataset will prove the effectiveness of the batch optimization method, which is quite efficient and accurate. In addition, for the local-level relative roto-translation estimation, yet-another robust wall-detection-based scan-matching algorithm is proposed and implemented to enhance the capability of the previous scan-matching algorithm: first, conditioned-Hough-Transform-and-linear-regression-based line-segment detection is performed to detect the line segments from the raw laser-scan-range data; second, wall detection is done to select the line segments that correspond to the walls of the environment; third, matching by fitting point to line is executed to estimate the roto-translation value. The experimental result will verify the effectiveness of this algorithm even when the moving object is close to the wall and there is much rotation error in the input odometry data. Moreover, on the global level, with the knowledge of the estimated global robot poses for the transformations between the local robot frames and the global inertial frame, the local line-feature maps can be transformed and integrated to the global frame in order to construct the global line-feature-based map. The experimental verification will prove the effectiveness of the complete graph- based robust and mapping mapping algorithm both in simulation and in actual large-dataset hardware experiment. In conclusion, one hierarchical robust graph-based localization and mapping algorithm is designed and implemented in this dissertation for dynamic indoor scenarios, which solves the problem of localization and mapping in a robust and effective way. One possible future direction for the research is to adapt this full off-line graph-based SLAM approach to the on-line version to deal with live sensor data.

En aquesta tesis s'ha desenvolupat un sistema de control capaç d'optimitzar el funcionament dels Reactors Discontinus Seqüencials dins el camp de l'eliminació de matèria orgànica i nitrogen de les aigües residuals. El sistema de control permet ajustar en línia la durada de les etapes de reacció a partir de mesures directes o indirectes de sondes. En una primera etapa de la tesis s'ha estudiat la calibració de models matemàtics que permeten realitzar fàcilment provatures de diferents estratègies de control. A partir de l'anàlisis de dades històriques s'han plantejat diferents opcions per controlar l'SBR i les més convenients s'han provat mitjançant simulació. Després d'assegurar l'èxit de l'estratègia de control mitjançant simulacions s'ha implementat en una planta semi-industrial. Finalment es planteja l'estructura d'uns sistema supervisor encarregat de controlar el funcionament de l'SBR no només a nivell de fases sinó també a nivell cicle.
In this Thesis a control system has been developed which permits optimizing the performance of the Sequencing Batch Reactors (SBR) within the field of organic matter and nitrogen removal from the wastewater. This control system is based on the on-line adjustment of the length of the reaction phases using directly or indirectly the data acquired from the sensors. In a first stage of the Thesis the calibration of the activated sludge models is studied what permits obtaining models for testing different operating and control strategies. From the analysis of historical data several options for controlling the SBR are obtained and most suitable is tested using a simulation approach. Afterwards, the control strategy is implemented in a semi-industrial plant obtaining promising results. Finally, a proposal for a supervisory control system is presented which can be in charge of controlling the performance of the SBR not only at a phase level but also at cycle level.

Le développement d’un procédé de polymérisation en émulsion est complexe de par la nature hétérogène de ce type de réaction. En outre, dans le cas de la polymérisation en émulsion du fluorure de vinylidène, la difficulté est d’autant plus accrue que le monomère est habituellement en phase gazeuse ou supercritique dans les conditions d’intérêt. Or la littérature manque d’informations concernant ce type de procédé de synthèse du PVDF sous une pression comprise entre 30 bar et 90 bar.Ainsi cette thèse a pour objectif de contribuer à une meilleure compréhension des mécanismes cinétiques et de stabilisation intervenants dans la polymérisation radicalaire en émulsion du VDF dans des conditions supercritiques et, plus particulièrement, de fournir des données expérimentales nécessaires à l’élaboration de futurs modèles.Avant même d’entreprendre les études expérimentales, cette thèse s’intéresse d’abord aux aspects d’installation et d’automatisation de l’unité de polymérisation ainsi qu’au démarrage et à l’optimisation du réacteur. Ensuite, plusieurs tests sont réalisés afin de comprendre certaines caractéristiques du latex produit ainsi que certaines propriétés du tensioactif fluoré. Une nouvelle méthode a spécialement été développée afin de suivre le phénomène de coagulation des particules de polymère.Finalement des réactions sont réalisées par lot et en semi-continu et une étude paramétrique des conditions opératoires et de la composition des réactifs est effectuée afin d’évaluer leur impact sur l’évolution de la polymérisation en émulsion. Notamment, le profil de vitesse de polymérisation est obtenu par calorimétrie, à partir d’une approche pratique fondée sur un estimateur d’état en cascade, ainsi que sur la mesure de la consommation de monomère, et sur l’analyse gravimétrique réalisée par prélèvement
The heterogeneous nature of the conventional emulsion polymerization can render the process quite complex. In the case of the emulsion polymerisation of vinylidene fluoride (VDF), the situation is more complicated than for the majority of industrial processes because the monomer is typically either a gas or a supercritical fluid under the polymerization conditions of interest. Given the relatively high pressure required for this process (30bar

Made available in DSpace on 2015-05-08T14:53:24Z (GMT). No. of bitstreams: 1 ArquivoTotal.pdf: 2670073 bytes, checksum: b32f39ed0b53334c57b727abcef4f5cd (MD5) Previous issue date: 2014-08-30
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES
The work presented in this thesis concerns the problem of sizing and sequencing batch production, single level, skilled (CLSP), in a manufacturing plant precast concrete used in construction, where the sizing and sequencing of lots production in a given period, represent a challenge for the company, which in turn seeks to minimize the costs of production, inventory, setup and improve the level of service customers' demands. The problem consists in defining the size of lots and production sequence of several blocks precast concrete in a press line, with setup times too high. It was proposed, therefore, a mathematical model whose objective is to minimize the costs of production, setup and inventory, involved in the manufacturing process of these products through the exact method of Integer Linear Programming. The presented model aims to solve the problem was proposed and applied to the real situation of the company for the purpose of comparison of current results with the results after application of the model .
O trabalho que se apresenta nesta dissertação diz respeito ao problema de dimensionamento e sequenciamento de lotes de produção, único nível, capacitado (CLSP), em uma fábrica de produtos pré-moldados de concreto usados na construção civil, onde o dimensionamento e sequenciamento dos lotes de produção num período pré-definido, representam um desafio para a empresa; que por sua vez busca minimizar os custos de produção, estoque, setup e melhorar o nível de atendimento das demandas dos clientes. O problema consiste em definir o tamanho dos lotes e sequência de produção de diversos blocos pré-moldados de concreto em uma linha de prensagem, com tempos de setup muito altos. Será proposto, portanto, um modelo matemático cujo objetivo é minimizar os custos de produção, setup e estoque, envolvidos no processo de fabricação destes produtos através do método exato de Programação Linear Inteira. O modelo a ser apresentado visa resolver o problema proposto e será aplicado em situação real da empresa para fins de comparação dos resultados atuais com os resultados após aplicação do modelo.

The Cal Poly Spacecraft Simulator is currently being developed for future testing and verifying theoretical control applications. This paper details the effort to balance the platform and remove undesired external torque from the system using System Identification technique developed by Patrick Healy. Since the relationship between the input and output of the system is linear, the least square method is proposed to identify the mass properties and location of center of mass of the system. The tests use four sine wave generators that are out of phase with different amplitudes as the inputs to excite various structural modes of the system. The outputs, angular rates of the platform, are measured by the newly implemented LN-200 Inertial Measurement Unit that helps reducing the measurement noise. Two test cases of 90o yaw rotations with the identified inertia were performed and validated against the computer simulation model; and the result shows that the test cases trajectories followed closely with the computer simulation model.

Les procédés électrochimiques occupent une part non-négligeable dans l'industrie chimique. Des études préliminaires montrent cependant que le mode de fonctionnement traditionnel de ces procédés (à potentiel ou à courant constant) n'est pas toujours le meilleur. Elles encouragent à envisager l'amélioration du fonctionnement des procédés électrochimiques par l'utilisation des méthodes d'optimisation dynamique. L’étude présentée dans cette thèse propose une méthodologie de l'optimisation dynamique et de la commande optimale dans le cas des procédés électrochimiques discontinus. On cherche ainsi à formuler et résoudre les problèmes d'optimisation du fonctionnement dynamique des réacteurs électrochimiques, à analyser les différentes sensibilités des paramètres du modèle et à appliquer les profils optimaux de commande en boucle fermée dans des conditions expérimentales aussi réalistes que possibles. La méthodologie est appliquée à différents problèmes électrochimiques tels que la minimisation de la consommation d'énergie électrique, la maximisation du rendement, de la sélectivité ou encore la minimisation du temps opératoire. Toutes les méthodes d'optimisation utilisées sont fondées sur le principe du maximum. Un inventaire non exhaustif des techniques de résolution de problèmes d'optimisation dynamique et leurs principales caractéristiques sont ainsi présentées. En intégrant au problème, des contraintes physiques, ces méthodes fournissent des conditions optimales de fonctionnement réalistes. Les performances de l'optimisation dynamique sont comparées à celles résultant des meilleures conditions de fonctionnement statiques. Les améliorations entre le mode de fonctionnement optimisé et le meilleur mode usuel peuvent s'élever à plusieurs dizaine de pourcent. Différentes formes de l'analyse de la sensibilité des paramètres du modèle soulignent la validité des résultats optimaux dans des conditions opératoires qui diffèrent de celles idéalement représentée dans le modèle des réacteurs considérés. Pour la mise en œuvre de la commande optimale, des techniques de reconstruction de l'état et une loi de commande en boucle fermée ont été mis au point.

Ce travail a eu pour objectif de déterminer les conditions optimales de production d’un caroténoïde original, la torularhodine, par Sporobolomyces ruberrimus, cultivée en réacteur discontinu. Cette souche est capable d’utiliser le glycérol technique comme source de carbone et d’énergie pour sa croissance et pour la production de caroténoïdes. D’abord, il s’est agi d’identifier les facteurs opératoires majeurs qui sont susceptibles d’avoir une influence sur la production de la torularhodine, au travers d’une étude préliminaire. L’identification expérimentale de ces facteurs d’action - la température, le taux d’oxygène dissous et la supplémentation en acide oléique - a été validée statistiquement, à des degrés divers, avant d’engager une étape d’optimisation par la construction d’un plan d’expériences multicritère. Celui-ci a conduit à l’établissement de modèles polynômiaux du second degré pour représenter l’effet conjugué des facteurs retenus et permettre la prédiction des valeurs de µmax et de concentration de torularhodine rapportée à la biomasse. Cette étude a alors été consacrée à un essai de fonctionnalisation de la torularhodine, à partir de sa fonction carboxylique, en vue de la stabilisation de la molécule dont l’activité antioxydante est élevée. L’acylation enzymatique de la lysine par la torularhodine a été envisagée. Les conditions d’acylation par la lipase B de C. antarctica ont été déterminées avec un caroténoïde modèle, la bixine. Le produit dérivé obtenu après transacylation a été purifié et a montré une activité antiradicalaire supérieure à celle de la bixine. Ces résultats permettent d’envisager la synthèse de peptides acylés avec ce type de caroténoïdes
The aim of this work was to determine the optimum of an original carotenoid, the torularhodin, produced by Sporobolomyces ruberrimus, in batch culture. A very interesting characteristic of this strain is its ability to consume raw glycerol as a carbon and energy source for microbial growth and carotenoid production. In the fist part of this study, the identification of operating parameters that have an influence on the optimum torularhodin production, was achieved. Experimental assays reinforced by a statistical study allowed to identify temperature, dissolved oxygen pressure and oleic acid supplementation, as the major parameters of influence, and then the integration of these data was performed for the construction of a multiobjective optimization based on a multicriteria experimental design. The establishment of a mathematical model of a second degree polynomial type was developed for the prediction of the values of µmax and of the torularhodin concentration reported to biomass. In the last part, considering that torularhodin has an important antioxidant property and it exhibits a free carboxyl acid function which can be used as acyl agent, a study of its structure modifying by an enzymatic way as a stabilization pattern was started. The experimental conditions of lysine acylation by the lipase B of Candida antarctica were determined using a model carotenoid, the bixin. The resulting product of the synthesis of bixin derivative was purified and showed an antiradical activity of 2.5 times higher than that of bixin. This result showed the ability of the acylation reaction of peptides with this kind of carotenoids

Doutoramento em Engenharia Química e Biológica
Os principais objectivos desta tese são: o desenho óptimo de experiências para a identificação de coeficientes de rendimento de um modelo não estruturado de um processo de fermentação semicontínua de Escherichia coli; a verificação experimental das trajectórias de alimentação obtidas por simulação; o desenvolvimento de estratégias de monitorização avançada para a estimação em linha de variáveis de estado e parâmetros cinéticos; e por fim o desenvolvimento de uma lei de controlo adaptativo para controlar a taxa específica de crescimento, com base em estratégias de alimentação de substrato com vista à maximização do crescimento e/ou produção. São apresentadas metodologias para o desenho óptimo de experiências, que visam a optimização da riqueza informativa das mesmas, quantificada por índices relativos à Matriz de Informação de Fisher. Embora, o modelo utilizado para descrever a fermentação semi-contínua de E. coli não esteja ainda optimizado em termos cinéticos e de algumas dificuldades encontradas na implementação prática dos resultados obtidos por simulação para o desenho óptimo de experiências, a qualidade da estimativa dos parâmetros, especialmente os do regime respirativo, é promissora. A incerteza das estimativas foi avaliada através de índices relacionados com o modelo de regressão linear múltipla, índices relativos à matriz de Fisher e pelo desenho das correspondentes elipses dos desvios. Os desvios associados a cada coeficiente mostram que ainda não foram encontrados os melhores valores. Procedeu-se também à investigação do papel do modelo dinâmico geral no desenho de sensores por programação. Foram aplicados três observadores – observador estendido de Kalman, observador assimptótico e observador por intervalo – para estimar a concentração de biomassa, tendo sido avaliado e comparado o seu desempenho bem como a sua flexibilidade. Os observadores estudados mostraram-se robustos, apresentando comportamentos complementares. Os observadores assimptóticos apresentam, em geral, um melhor desempenho que os observadores estendidos de Kalman. Os observadores por intervalo apresentam vantagens em termos de implementação prática, apresentando-se bastante promissores embora a sua validação experimental seja necessária. É apresentada uma lei de controlo adaptativo com modelo de referência que se traduz num controlo por antecipação/retroacção cuja acção de retroacção é do tipo PI, para controlar a taxa específica de crescimento. A robustez do algoritmo de controlo foi estudada por simulação numérica gerando dados “pseudo reais”, por aplicação de um ruído branco às variáveis medidas em linha, por alteração do valor de referência, por alteração do valor da concentração da glucose na alimentação e variando os valores nominais dos parâmetros do modelo. O estudo realizado permite concluir que a resposta do controlador é em geral satisfatória, sendo capaz de manter o valor da taxa específica de crescimento na vizinhança do valor de referência pretendido e inferior a um valor que conduz à formação de acetato, revestindo-se este facto de grande importância numa situação real, em especial, numa fermentação cujo objectivo seja a produção, nomeadamente de proteínas recombinadas. Foram ainda, analisados diferentes métodos de sintonização dos parâmetros do controlador, podendo concluir-se que, em geral, o método de sintonização automática com recurso à regra de adaptação dos parâmetros em função do erro relativo do controlador foi o que apresentou um melhor desempenho global. Este mecanismo de sintonização automática demonstrou capacidade para melhorar o desempenho do controlador ajustando continuamente os seus parâmetros.
The main objectives of this thesis are: the optimal experiment design for yield coefficients estimation in an unstructured growth model for Escherichia coli fed-batch fermentation; the experimental validation of the simulated feed trajectories; the development of advanced monitoring strategies for the on-line estimation of state variables and kinetic parameters; and at last the development of an adaptive control law, based on optimal substrate feed strategies in order to increase the growth and/or the production. Methodologies for the optimal experimental design are presented, in order to optimise the richness of data coming out from experiments, quantified by indexes based on the Fisher Information Matrix. Although the model used to describe the E. coli fed-batch fermentation is not optimised from the kinetic properties point of view and the fact that some difficulties were encountered in practical implementation of the simulated results obtained with the optimal experimental design, the estimated parameter quality, especially for the oxidative regimen, is promising. The estimation uncertainty was evaluated by means of indexes related with multiple linear regression model, indexes related to the Fisher matrix as well as by the construction of the related deviation ellipses. The deviations associated to each coefficient show that the best values were not yet found. The role of the general dynamical model was also investigated in which concerns the design of state observers, also called software sensors. The performance of three observer classes was compared: Kalman extended observer, assimptotic observer and interval observer. The studied observers showed good performance and robustness, being complementary of each other. Assimptotic observers showed, in general, a better performance than the Kalman extended observer. Interval observers presented advantages concerning practical implementation, showing a promising behaviour although experimental validation is needed. A model reference adaptive control law is presented and can be interpreted as a PI like feedforward/feedback controller, for specific growth rate control. Algorithm robustness was studied using “pseudo real” data obtained by numerical simulation, by applying a white noise to the on-line measured variables, by modifying the set-point value, by changing the glucose concentration value of the feed rate and varying the nominal model parameter value. The study made allowed to conclude that the controller response is, generally, satisfactory being able to keep the specific growth rate value in the proximity of the desired set-point and lower than the value that permits acetate formation, which is of major importance namely for real cases, specially, in a fermentation which objective was the production of recombinant proteins. Different tuning devices for controller parameters were analysed being the better performance achieved by the automatic tuning method with an adaptation rate as a function of the controller relative error. This automatic tuning mechanism was able to improve the controller performance adjusting continuously its parameters.

Made available in DSpace on 2015-05-14T13:21:37Z (GMT). No. of bitstreams: 1 parte1.pdf: 2426440 bytes, checksum: ed3c18b465357e2c561cffbd380b983f (MD5) Previous issue date: 2008-07-25
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES
The chemiluminescence (CL) of luminol-cobalt(II) reaction has been used in analytical systems for capillary electrophoresis, chromatography, -TAS (microTotal Analysis System), etc. Even with the diversity of applications, the steps of this reaction have not been completely elucidated and have been still a subject of research and controversy. Many works reported only to the isolated studies of the influence of each variable in the luminol-cobalt(II) reaction and, as such studies do not consider the interactions among variables, they do not exploit the total performance of this reaction in order to increase the chemiluminescent intensity (ICL). In this context, the use of an experimental design and a ChemiLumimetric Flow-Batch System (CLFBS) is proposed in this work in order to study and to optimize this reaction, aiming at to develop an automatic method to determine vitamin B12 (VB12) in drugs. For optimization purpose, a 24 factorial design was carried out fixing the cobalt concentration at 3.0 g L-1 and varying the concentration of luminol, hydrogen peroxide, sodium hydroxide and the order of mixture of these reagents. Sixteen assays were performed in five replicates, generating a total of eighty experiments. The analysis using normal plot of the experimental design revealed that the luminol concentration and the order of mixture of the reagents are variables more important than the NaOH or H2O2 concentrations to luminol-cobalt(II) reaction. These two parameters were responsible to enhance the chemiluminescent signal in about 80%. Another study was carried out in order to evaluate the CLFBS performance by using the optimized variables which were suggested by the factorial design study. Calibration curves were built by using standard solution of Co(II) and VB12 and the analytical parameters for Co(II) curve were: ΔICL = -21.39 +1771.37[Co2+] (r2 = 0.9996), LD and LQ = 12.0 ng L-1, RSD = 1.8% (n = 5), analytical sensitivity = 1947.29 W/g L-1; and VB12 curve were: ΔICL = -186.71 + 12.90 [VB12] (r2 = 0.9999), LD = 14.53 mg L-1 and LQ = 14.70 mg L-1, RSD = 2.1% (n = 4) and analytical sensitivity = 10.76 W/μg L-1. The results of the vitamin B12 analysis in drug samples employing the luminol-cobalt(II) reaction and CLFBS were enough satisfactory. Relative errors smaller than 4% were obtained by using curve calibration or standard addition method. The recovery studies yield very good values, which were of 97 to 103%. In addition, a good agreement was obtained when a drug sample was analyzed by the proposed and the reference (HPLC) method. Thus, the automatic chemilumimetric method, which was here developed and optimized, can be considered a promising alternative to quality control of vitamin B12 in drugs.
A quimiluminescência (QL) da reação luminol-cobalto(II) vem sendo empregada em sistemas analíticos de eletroforese capilar, cromatografia, -TAS (microTotal Analysis System), etc. Mesmo com a diversidade de aplicações, essa reação possui etapas pouco elucidadas que continuam sendo alvo de pesquisas e polêmicas. Muitos trabalhos relatam apenas estudos isolados da influência de cada variável na reação luminol-cobalto(II) e, por não considerar as interações entre as variáveis, estes estudos não exploram ao máximo a performance analítica desta reação. Neste contexto, foi proposto neste trabalho o uso de um planejamento experimental e de um Sistema Quimilumimétrico Fluxo-Batelada (SQLFB) para o estudo e a otimização desta reação com vista a desenvolver um método automático para a determinação de vitamina B12 (VB12) em medicamentos. Na otimização, foi utilizado um planejamento fatorial 24 em que foi mantido constante a concentração de cobalto(II) em 3,0 g L-1 enquanto variou-se as concentrações de luminol, peróxido de hidrogênio, hidróxido de sódio e a ordem de mistura destes constituintes. Foram efetuados dezesseis ensaios em quintuplicata, perfazendo um total de 80 experimentos realizados. A análise usando o gráfico normal do planejamento experimental revelou que a concentração de luminol ([Lu]) e a ordem da mistura (OM) dos reagentes são variáveis mais importante do que a concentração de NaOH e H2O2 para a reação luminol-cobalto(II). Ambas variáveis [Lu] e OM foram responsáveis por aumentar o sinal quimiluminescente em cerca de 80%. Um outro estudo foi realizado para avaliar a performance do SQLFB, empregando as variáveis otimizadas que foram sugeridas pelo estudo do planejamento fatorial. Curvas de calibração foram construídas utilizando soluções padrão de Co(II) e de VB12 e os parâmetros analíticos para a curva Co(II) foram: ΔIQL=-21,39+1771,37[Co2+](r2 = 0,9996), LD=1,54 ng L-1 e LQ=5,13 ng L-1, DPR=1,8% (n=5) e sensibilidade analítica=1947,29 Watts/μg L-1; e para a curva VB12 foram: ΔIQL= -186,71 + 12,90[VB12] (r2 = 0,9999), LD=0,89 μg L-1 e LQ = 2,98 μg L-1, DPR=2,1% (n=4) e sensibilidade analítica=10,76 Watts/μg L-1. Os resultados obtidos nas análises de vitamina B12 em medicamentos empregando a reação luminol-cobalto(II) e o SQLFB foram bastante satisfatórios. Erros relativos menores do que 4% foram obtidos empregando as técnicas de curva de calibração e de adição de padrão. Em estudos de recuperação, os valores foram também muito bons, ficando entre 97 e 103%. Além disso, uma boa concordância entre os resultados foi obtida quando uma amostra foi analisada empregando o método aqui proposto e o método de referência (HPLC). Portanto, o método quimilumimétrico automático aqui desenvolvido e otimizado pode ser considerado uma alternativa promissora para o controle de qualidade de vitamina B12 em medicamentos.

Os principais objectivos desta tese são: o desenho óptimo de experiências para a identificação de coeficientes de rendimento de um modelo não estruturado de um processo de fermentação semicontínua de Escherichia coli; a verificação experimental das trajectórias de alimentação obtidas por simulação; o desenvolvimento de estratégias de monitorização avançada para a estimação em linha de variáveis de estado e parâmetros cinéticos; e por fim o desenvolvimento de uma lei de controlo adaptativo para controlar a taxa específica de crescimento, com base em estratégias de alimentação de substrato com vista à maximização do crescimento e/ou produção. São apresentadas metodologias para o desenho óptimo de experiências, que visam a optimização da riqueza informativa das mesmas, quantificada por índices relativos à Matriz de Informação de Fisher. Embora, o modelo utilizado para descrever a fermentação semi-contínua de E. coli não esteja ainda optimizado em termos cinéticos e de algumas dificuldades encontradas na implementação prática dos resultados obtidos por simulação para o desenho óptimo de experiências, a qualidade da estimativa dos parâmetros, especialmente os do regime respirativo, é promissora. A incerteza das estimativas foi avaliada através de índices relacionados com o modelo de regressão linear múltipla, índices relativos à matriz de Fisher e pelo desenho das correspondentes elipses dos desvios. Os desvios associados a cada coeficiente mostram que ainda não foram encontrados os melhores valores. Procedeu-se também à investigação do papel do modelo dinâmico geral no desenho de sensores por programação. Foram aplicados três observadores – observador estendido de Kalman, observador assimptótico e observador por intervalo – para estimar a concentração de biomassa, tendo sido avaliado e comparado o seu desempenho bem como a sua flexibilidade. Os observadores estudados mostraram-se robustos, apresentando comportamentos complementares. Os observadores assimptóticos apresentam, em geral, um melhor desempenho que os observadores estendidos de Kalman. Os observadores por intervalo apresentam vantagens em termos de implementação prática, apresentando-se bastante promissores embora a sua validação experimental seja necessária. É apresentada uma lei de controlo adaptativo com modelo de referência que se traduz num controlo por antecipação/retroacção cuja acção de retroacção é do tipo PI, para controlar a taxa específica de crescimento. A robustez do algoritmo de controlo foi estudada por simulação numérica gerando dados “pseudo reais”, por aplicação de um ruído branco às variáveis medidas em linha, por alteração do valor de referência, por alteração do valor da concentração da glucose na alimentação e variando os valores nominais dos parâmetros do modelo. O estudo realizado permite concluir que a resposta do controlador é em geral satisfatória, sendo capaz de manter o valor da taxa específica de crescimento na vizinhança do valor de referência pretendido e inferior a um valor que conduz à formação de acetato, revestindo-se este facto de grande importância numa situação real, em especial, numa fermentação cujo objectivo seja a produção, nomeadamente de proteínas recombinadas. Foram ainda, analisados diferentes métodos de sintonização dos parâmetros do controlador, podendo concluir-se que, em geral, o método de sintonização automática com recurso à regra de adaptação dos parâmetros em função do erro relativo do controlador foi o que apresentou um melhor desempenho global. Este mecanismo de sintonização automática demonstrou capacidade para melhorar o desempenho do controlador ajustando continuamente os seus parâmetros.
The main objectives of this thesis are: the optimal experiment design for yield coefficients estimation in an unstructured growth model for Escherichia coli fed-batch fermentation; the experimental validation of the simulated feed trajectories; the development of advanced monitoring strategies for the on-line estimation of state variables and kinetic parameters; and at last the development of an adaptive control law, based on optimal substrate feed strategies in order to increase the growth and/or the production. Methodologies for the optimal experimental design are presented, in order to optimise the richness of data coming out from experiments, quantified by indexes based on the Fisher Information Matrix. Although the model used to describe the E. coli fed-batch fermentation is not optimised from the kinetic properties point of view and the fact that some difficulties were encountered in practical implementation of the simulated results obtained with the optimal experimental design, the estimated parameter quality, especially for the oxidative regimen, is promising. The estimation uncertainty was evaluated by means of indexes related with multiple linear regression model, indexes related to the Fisher matrix as well as by the construction of the related deviation ellipses. The deviations associated to each coefficient show that the best values were not yet found. The role of the general dynamical model was also investigated in which concerns the design of state observers, also called software sensors. The performance of three observer classes was compared: Kalman extended observer, assimptotic observer and interval observer. The studied observers showed good performance and robustness, being complementary of each other. Assimptotic observers showed, in general, a better performance than the Kalman extended observer. Interval observers presented advantages concerning practical implementation, showing a promising behaviour although experimental validation is needed. A model reference adaptive control law is presented and can be interpreted as a PI like feedforward/feedback controller, for specific growth rate control. Algorithm robustness was studied using “pseudo real” data obtained by numerical simulation, by applying a white noise to the on-line measured variables, by modifying the set-point value, by changing the glucose concentration value of the feed rate and varying the nominal model parameter value. The study made allowed to conclude that the controller response is, generally, satisfactory being able to keep the specific growth rate value in the proximity of the desired set-point and lower than the value that permits acetate formation, which is of major importance namely for real cases, specially, in a fermentation which objective was the production of recombinant proteins. Different tuning devices for controller parameters were analysed being the better performance achieved by the automatic tuning method with an adaptation rate as a function of the controller relative error. This automatic tuning mechanism was able to improve the controller performance adjusting continuously its parameters.

no
Batch and semi-batch reactors are widely used for fine chemical productions. The target in the fine chemical industry is to produce a high quality product and operational optimization is the key-element to match it. This work investigates how batch and semi-batch reactors can be optimized in order to increase the yield of a desired product. Optimization problem is formulated and applied to calculate the optimal operating parameters such as the reactor temperature and the feed flow rate. Comparison and considerations on the two reactor configurations are given.

碩士
國立臺灣大學
化學工程學研究所
104
Three published crystallization kinetics from batch crystallizer is compiled, and the simulation and optimization of crystallization process are also provided. The results suggest that a late growth policy would be appropriate to minimize the final nucleated crystal mass. However determining the optimal curve of temperature or solvent mass is difficult and requires a kinetic model. Adding constraints on operating curve is a common method in industries. The simulation results of optimal procedure with constraints illustrate a trade-off between mass of nucleated crystals and equipment costs. Another problem often encountered in practical operation is hard to collecting useful data to find crystal kinetics. When dimensionless temperature or solvent mass is plotted versus dimensionless time, the optimal trajectory for many systems is similar for a given constraint. Therefore, for each constraint the results of many systems are fitted to a single curve which can be applied without knowledge of the kinetic parameters. Knowing the total batch time, initial and final value of manipulated variable, we can find a nearly optimal operating curve in a specific constraint. The average deviation between each optimal trajectory and regressed curve is 5.03%. Mullin-nyvlt, cubic trajectory and regressed curve may applicable in different system. Linear trajectory is no applicable in any system.

碩士
國立成功大學
化學工程學系碩博士班
95
A　general　mixed-integer　nonlinear　programming　model　(MINLP)　is　developed　in　this　study　to　synthesize　water　networks　in　batch　processes.　The　proposed　model　formulation　is　believed　to　be　superior　to　the　available　ones.　In　the　past,　the　tasks　of　optimizing　batch　schedules,　process　water　reuse　subsystems　and　wastewater　treatment　subsystems　were　performed　individually.　In　this　study,　all　three　optimization　problems　are　incorporated　in　the　same　mathematical　programming　model.　By　properly　addressing　the　issue　of　interaction　between　subsystems,　better　overall　designs　can　be　generated.　The　resulting　design　specifications　include:　the　short-term　production　schedule,　the　number　and　sizes　of　buffer　tanks,　the　physical　configuration　of　pipeline　network,　and　the　operating　policies　of　water　flows.　The　network　structure　can　also　be　strategically　manipulated　by　imposing　suitable　logic　constraints.　A　series　of　illustrative　examples　are　presented　to　demonstrate　the　effectiveness　of　the　proposed　approach.

碩士
國立中正大學
化學工程研究所
84
Until lately, almost all control problems in process engineering have been considered as single objective problems. In addition to economic efficiency, the performance of many process systems requires an evaluation through variety criteria. Environmental quality, reliability, safety, etc., have become as important as their economic efficiency. Since different perfor-mance objectives may compete against each other, the single objective approach can only give a compromised solution to the overall problem but does not provide more insight into the conflicts among the competing objectives. Pareto, or non-inferior sets occur when the objectives are in conflicts and no objective can be improved without sacrifice of the others. From above mentioned, in order to determine the trade-off information between the different goals, multiobjective optimization methods have to beemployed in which two or more non-commensurable objectives are consi-dered simultaneously.The major purpose of this paper is to extend a method of iterative dynamic programming（IDP）for the application of multiobjective opti -mal problems. By using theε- constraint approach, such problems are firstconverted into a single objective optimal control problem with constraints. This single objective problem could be then solved by conventional IDP incorporated with the penalty functions. However, it is impossible to obtain trade-off rates between multiobjective functions for lack of the ability to estimate the Lagrange multipliers. In this study, the augmented Lagrange multiplier updating method embedding in IDP is introduced to estimate Lagrange multipliers for the optimal problems. This method besides affords an option to decide the initial choices of the penalty parameters. Two batch chemical processes, bulk polymerization of styrene and crystallization of sugar are simulated to investigate the effect of using the multiplier updating method. In this paper, a better quantitative understanding of the trade-offs between the objectives has been achieved by relating the trade-off rates within the Pareto solutions to the Lagrange multipliers.

Thesis (Ph. D.)--Florida State University, 2004.
Advisor: Dr. Srinivas Palanki, Florida State University, College of Engineering, Dept. of Chemical Engineering. Title and description from dissertation home page (viewed Oct. 06, 2004). Includes bibliographical references.