Dissertations / Theses on the topic 'On-line process control'
To see the other types of publications on this topic, follow the link: On-line process control.
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'On-line process control.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
1
Oisiovici, Ronia Marques. "Controle "on-line" de colunas de destilação em batelada de alta pureza." [s.n.], 2001. http://repositorio.unicamp.br/jspui/handle/REPOSIP/267512.
Full textAbstract:
Orientadores: Sandra Lucia da Cruz, João Alexandre F. R. PereiraTese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Quimica
Made available in DSpace on 2018-07-27T15:50:28Z (GMT). No. of bitstreams: 1 Oisiovici_RoniaMarques_D.pdf: 4098399 bytes, checksum: e53dcc32c80f82bc98dfe9a977ac5466 (MD5) Previous issue date: 2001
Resumo: Os processos em batelada são largamente empregados nas indústrias de Química Fina, produtos bioquímicos, polímeros e fármacos. As plantas em batelada tendem a crescer, já que esse tipo de configuração apresenta características que são desejáveis nas chamadas "plantas do futuro": flexibilidade operacional, resposta rápida às exigências do mercado, capacidade de processar produtos de alta pureza e alto valor agregado. A destilação em batelada, em particular, é a operação unitária mais freqüentemente utilizada nas plantas em batelada. Algumas dificuldades encontradas na sua automação são: fortes não-linearidades, desvios entre o modelo e o processo, ganhos variáveis no tempo, presença de ruídos nas medidas, analisadores de composição "on-line" complexos ou lentos, estimativa do estado do sistema dificil e cara em termos computacionais. O controle de colunas de destilação contínuas de alta pureza é um tópico bastante abordado na literatura. Porém, trabalhos sobre destilação em batelada de alta pureza são escassos e, normalmente, apenas simulações são apresentadas. Esse trabalho teve como objetivo o desenvolvimento de um sistema de controle por inferência para colunas de destilação em batelada de alta pureza robusto às dificuldades de automação encontradas na prática. O algoritmo de controle foi testado por simulação e, para confirmar a aplicabilidade do controlador desenvolvido, corridas experimentais foram realizadas numa coluna de destilação em batelada em escala piloto. Um Filtro de Kalman Estendido (FKE) para colunas de destilação em batelada foi desenvolvido para fornecer estimativas de composições instantâneas a partir de algumas medidas de temperatura. O FKE foi combinado a uma estratégia de controle baseada na estrutura GLC ("Globa1ly Linearizing Control"), resultando num sistema de controle por inferência que mantém constante a composição do produto de topo de colunas de destilação em batelada de alta pureza. O controlador GLC foi testado numa coluna de destilação em batelada em escala piloto com 29 pratos perfurados. As corridas experimentais foram realizadas com o sistema etanol/l-propanol. A interface entre o computador e o processo foi feita através de uma placa ADIDA. Termopares foram utilizados para medir temperaturas instantâneas em alguns estágios da coluna. A cada período de amostragem, o FKE fornece valores de composição instantâneos, utilizando os dados de temperatura lidos pelo computador. Com o objetivo de manter a pureza do produto de topo no valor especificado, o controlador GLC atualiza a razão de refluxo baseado nas estimativas mais recentes de composição. A razão de refluxo calculada é, então, implementada pelo computador que controla a posição e o tempo de abertura da válvula magnética de refluxo. O número de sensores e o intervalo de integração do modelo da coluna foram variáveis importantes no projeto do FKE. Convergência mais rápida e estimativas mais exatas foram obtidas aumentando-se o número de sensores e/ou reduzindo-se o intervalo de integração. No entanto, acima de um certo número de sensores, as melhorias na exatidão das estimativas são insignificantes. Nas corridas em malha fechada, a variável manipulada (razão de refluxo) apresentou um comportamento do tipo liga/desliga quando os sensores de temperatura foram posicionados nos estágios próximos ao topo da coluna, onde as variações de temperatura são pequenas. Afastando-se os sensores do topo da coluna, perfis mais suaves das variáveis manipulada e controlada foram observados e desvios inferiores a 0,2% entre a composição média do destilado estimada e a composição média do destilado medida foram obtidos. Para o controle da composição do destilado de alta pureza, os sensores devem ficar afastados do topo, mas não tão distantes a ponto de diminuir a exatidão das estimativas de composição. Devido à sua robustez em relação aos desvios entre o modelo e o processo, incerteza nas condições iniciais da coluna e a presença de ruídos nas medidas de temperatura, o controlador por inferência desenvolvido nesse trabalho, além de ser flexível, mostrou-se eficaz em aplicações práticas
Abstract: Batch processing has been widely used in the production of fine chemicaIs, biochemicals, polymers and phannaceuticals. Batch configurations present features which are desirable in the ca1led "future plants": flexibility of operation, rapid response to changing market demands, suitability for manufacturing high-quality and high value-added products. Batch distillation is the most frequent separation method in batch processes. Some ofthe challenges to be faced in the automation ofbatch plants are: pronounced nonHnearities, process/model mismatch, time varying process gains, presence of measurement noise, on-line sensors are unavailable or give delayed results, state estimation is difficult and computationally expensive. While the control of high-purity continuous distillation column is a subject often addressed in literature, there are very few papers regarding the control of high-purity batch distillation units. This work aimed to develop an inferential control system for high-purity batch distillation colurnns, which is able to cape with the automation difficulties common1yencountered in practice. The control algorithm was first tested by simulation and then experimental runs were carried out in a pilot-scale batch distillation column to confirm the practical appHcability ofthe designed inferential control system. An Extended Kalman Filter (EKF) for batch distillation columns was developed to provide fast and reHable instantaneous composition values from few temperature measurements. The EKF was then combined with a control strategy based on the Globally Linearizing Control (GLC) structure, resulting in an inferential control system for constant distillate operation of high-purity batch distillation columns. The GLC controller was tested in a pilot-scale batch distillation column with 29 sieve trays. The separation of ethanol/l-propanol mixtures were considered in the experimental runs. The interface between the computer and the process was achieved by an AD/DA converter. Thennocouples were used to measure instantaneous temperature values at some column stages. At each sampling period, the EKF provides instantaneous composition values from the temperature data the computer acquires. In order to keep the product quality at the set-point, the GLC controller updates the reflux ratio using the composition estimates, and the computer controls the position ofthe reflux magnetic valve to implement the calculated reflux ratio. The number of sensors and the model integration interval showed to be important variables in the design ofthe EKF. Faster convergence and more accurate estimates were obtained by increasing the number of sensors and/or reducing the integration intervalo However, above a certain number of sensors the improvement in the EKF performance may not be significant. Spiky reflux ratio profiles were observed when sensors were placed next to the top stages, where the temperature variations are small. As the sensors were placed further from the top stages, smoother manipulated and controlled variable profiles were obtained and the distillate product met the specified purity. The inferential control system achieved tight composition control. The selection of the sensors locations must be a trade-off between obtaining noise reduction and guaranteeing that the EKF will provide acceptable estimate accuracy. Due to its robustness with regards to plant model mismatch, uncertainty in the initial system state and measurement errors, the proposed inferential control scheme is flexible and demonstrated to be feasible for practical on-line applications
Doutorado
Sistemas de Processos Quimicos e Informatica
Doutor em Engenharia Química
2
Dowey, Stephen James. "Advanced on-line and off-line process control for surface-engineered applications." Thesis, University of Hull, 1999. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.310319.
Full text
3
Myerscough, James John. "On-line geometric control of a spray deposition process." Thesis, University of Oxford, 1992. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.320688.
Full text
4
Belger, I. S. "Real-time on-line control of a fermentation process." Thesis, Liverpool John Moores University, 1991. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.292336.
Full text
5
Guh, Ruey-Shiang. "On-line statistical process control : a hybrid intelligent approach." Thesis, University of Nottingham, 2000. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.311818.
Full text
6
Ha, Sungdo. "On-line control of process uniformity using categorized variabilities." Thesis, Massachusetts Institute of Technology, 1993. http://hdl.handle.net/1721.1/12335.
Full text
7
Montgomery, Paul Anthony. "On-line optimisation of a multivariable fed-batch fermentation process." Thesis, Liverpool John Moores University, 1989. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.252788.
Full text
8
Zhang, Jie. "Expert systems in on-line process control and fault diagnosis." Thesis, City University London, 1991. http://openaccess.city.ac.uk/7682/.
Full textAbstract:
In this research expert systems for on-line process control and fault diagnosis have been investigated and the majority of the research is on using expert systems in on-line process fault diagnosis. Several on-line expert systems, including a rule based controller and several fault diagnosis systems, have been developed in this research and are reported in this thesis. The research results obtained demonstrate that rule based controllers can be used in situations where mathematical models for the controlled process cannot be obtained or are very difficult to obtain. The research on on-line fault diagnosis emphasises deep knowledge based approaches. Two avenues in deep knowledge based approaches, namely causal search and qualitative modelling based diagnosis, have been investigated. In the approach of causal search the research results reveal that diagnostic efficiency can be achieved through structural decomposition. A systematic approach for developing diagnostic rules based on structural decomposition is presented in this thesis. Much of the research has been done on qualitative model based fault diagnosis. A qualitative reasoning method which utilizes knowledge on the quantitative relations among variables to reduce ambiguity and can cope with a wider range of situations than Raiman's Order of Magnitude Reasoning is investigated. In the qualitative model based diagnosis the function of the qualitative model is to predict the behaviour of the process under various hypotheses and, therefore, to verify these hypotheses. Further research concerning self-reasoning has been done for the qualitative model based diagnosis approach. Self-reasoning is achieved by backward tracing through the model of the diagnosis system and makes this diagnosis system more intelligent. Self-learning of heuristic rules based on qualitative modelling is investigated and heuristic rules can add efficiency to model based diagnosis. During investigating self-learning of heuristic rules, the good learning property of neural networks is recognised and, neural networks based on-line fault diagnoses are also investigated. The research results reveal that neural networks based diagnosis systems are easy to develop and perform robustly provided that the training data are available.
9
Fronimidis, Dimitris. "On-line monitoring and control of the injection moulding process." Thesis, University of Bath, 2007. https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.436877.
Full text
10
Naghdy, G. "Identification, modelling and on-line computer control of a wastewater treatment process." Thesis, University of Portsmouth, 1986. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.370913.
Full text
11
Meadowcroft, Simon Richard. "The application of on-line expert systems in the control of pigment kilns." Thesis, University of Huddersfield, 1991. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.305046.
Full text
12
Beaupré, Mathieu. "Characterization of on-line sensors for water quality monitoring and process control." Thesis, Université Laval, 2010. http://www.theses.ulaval.ca/2010/26963/26963.pdf.
Full text
13
Oztemel, Ercan. "Integrating expert systems and neural networks for on-line statistical process control." Thesis, Cardiff University, 1992. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.307382.
Full text
14
Abdullah, N. "Augmented integrated system optimisation and parameter estimation techniques for on-line hierarchical control of large scale industrial processes." Thesis, City University London, 1988. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.382827.
Full text
15
Soderstrom, Tyler Andrew. "Integration of on-line data reconciliation and bias identification techniques /." Full text (PDF) from UMI/Dissertation Abstracts International, 2001. http://wwwlib.umi.com/cr/utexas/fullcit?p3008445.
Full text
16
Porrino, Alessandre. "Statistical analysis of deterministic textures in steel sheet production." Thesis, Brunel University, 2004. http://bura.brunel.ac.uk/handle/2438/4980.
Full textAbstract:
Textured surfaces are universally adopted in the steel sheet production industry, and manufacturers are continuously improving the quality of the finished products through intense research in the surface characterisation field. Deterministic Surfaces are textured with specifically designed rolls in order to present a certain degree of regularity, which allows better control over the functional behaviour of the metal sheets. The regularity of the texture impressed on the steel sheets also allows unconventional approaches to surface characterisation and to the assessment of the texture's structure. Statistical analysis is the most effective way to target the isolation of the deterministic part of the surface, which represents the desired product, from the stochastic part, called ‘noise’ and associated with the inaccuracies of production and measurement. This work addresses the problem of characterisation of deterministic textures through statistical analysis, proposing innovative filtering techniques aimed at the realisation of an On-line Process Control System. Firstly the techniques proposed are theoretically formulated and studied, addressing in particular the physical meaning of the geometrical parameters extracted through statistical analysis of highly correlated portions of the textures. A method for isolating the deterministic textures present on a surface, called the Statistical Surface Filter, is presented and discussed in detail, and tested on existing laboratory samples. Secondly the filter is applied to preliminary measurements acquired by an innovative on-line measurement system currently under development, and evidence is shown that the technique is effective in separating the information regarding the regular patterns from the stochastic noise. The possible applications to on-line Statistical Process Control are discussed. Thirdly, the Statistical Surface Filter is tested on a set of measurements representing texturing rolls and textured sheets with different characteristics; statistical analysis of the surface parameters extracted from the filtered surfaces show that the technique allows the assessment of the different contributions of the various stages of the texturing process to the final product. Finally, a software package is implemented for the practical application of the filtering techniques and the parameters extraction; the algorithms that perform the statistical filtering are described and discussed, concluding with the operations of optimisation and fine-tuning for production-line implementation.
17
Yang, Bing. "Non-isothermal flow of polymer melt and on-line computer control of profile extrusion using thermal method /." The Ohio State University, 1987. http://rave.ohiolink.edu/etdc/view?acc_num=osu1487323583623366.
Full text
18
Wadewitz, Thomas R. (Thomas Richard). "Analysis of investments in process control tools : laser scanner used for on-line inspection in a polyester sheet production process." Thesis, Massachusetts Institute of Technology, 1995. http://hdl.handle.net/1721.1/32173.
Full textAbstract:
Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 1995, and Thesis (M.S.)--Massachusetts Institute of Technology, Sloan School of Management, 1995.Includes bibliographical references (p. 80).
by Thomas R. Wadewitz.
M.S.
19
Andersson, Niklas. "Jämförelse av On-line med off-line mätning för ökad processkontroll av kemikaliedoseringar vid kartongtillverkning." Thesis, Örebro universitet, Institutionen för naturvetenskap och teknik, 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:oru:diva-69763.
Full textAbstract:
"Vi utmanar konventionella förpackningar för en hållbar framtid." BillerudKorsnäs AB bildades år 2012 och inriktar sig mot pappersindustrin. På bruket i Frövi tillverkas vätskekartong för mejeriprodukter samt förpackningskartong, vilket görs med en produktionskapacitet av 470 000 ton/år. För att skapa god kartongkvalitét krävs rätt kemikaliedosering, vilket säkerställs genom provtagning. Genom att investera i ny mätutrustning ges därmed möjligheten till en jämnare mätprocess och därmed bättre styrmöjlighet. Detta examensarbete syftar till jämförelse av mätutrustning för jämnare kemikaliedosering. I denna rapport görs jämförelse av off-och online-utrustning och rutiner vid mätning på olika kartongskikt för att därefter avslutningsvis bedöma resultat och fortsatt arbete för en jämnare kemikaliedosering. Dessa jämförelser påvisar en titreringslösning med för hög koncentration för ett skikt vilket innebär osäkra mätresultat. Detta innebär att titratlösning av lägre koncentration måste komplettera den nuvarande för säkrare mätresultat."We challenge conventional packaging for a sustainable future" BillerudKorsnäs AB was formed in 2012 and focuses on the paper industry. At the mill in Frövi they create liquid cartonboard for dairy products and packing cartonboard with a production capacity of 470 000 tonnes/year. For making good quality possible proper chemical dosage is required and to guarantee that water tests has to be made. Investing in new measuring equipment will allow steady measurement process. This project is comparing measurement equipment for a steady chemical dosage. In this report comparing between off-line and on-line measurement equipments and routines when measuring different cartonboard layers. This will help the company in there further work with the process for a steady chemical dosage. The result of the comparisons demonstrate a high concentration of titration solution for one layer which contributes to uncertain results. This means that the company have to use a titration solution of lower concentration as well as the present obtain reliable results.
20
Singh, Rahul. "A model to integrate Data Mining and On-line Analytical Processing: with application to Real Time Process Control." VCU Scholars Compass, 1999. https://scholarscompass.vcu.edu/etd/5521.
Full textAbstract:
Since the widespread use of computers in business and industry, a lot of research has been done on the design of computer systems to support the decision making task. Decision support systems support decision makers in solving unstructured decision problems by providing tools to help understand and analyze decision problems to help make better decisions. Artificial intelligence is concerned with creating computer systems that perform tasks that would require intelligence if performed by humans. Much research has focused on using artificial intelligence to develop decision support systems to provide intelligent decision support.
Knowledge discovery from databases, centers around data mining algorithms to discover novel and potentially useful information contained in the large volumes of data that is ubiquitous in contemporary business organizations. Data mining deals with large volumes of data and tries to develop multiple views that the decision maker can use to study this multi-dimensional data. On-line analytical processing (OLAP) provides a mechanism that supports multiple views of multi-dimensional data to facilitate efficient analysis. These two techniques together can provide a powerful mechanism for the analysis of large quantities of data to aid the task of making decisions.
This research develops a model for the real time process control of a large manufacturing process using an integrated approach of data mining and on-line analytical processing. Data mining is used to develop models of the process based on the large volumes of the process data. The purpose is to provide prediction and explanatory capability based on the models of the data and to allow for efficient generation of multiple views of the data so as to support analysis on multiple levels. Artificial neural networks provide a mechanism for predicting the behavior of nonlinear systems, while decision trees provide a mechanism for the explanation of states of systems given a set of inputs and outputs. OLAP is used to generate multidimensional views of the data and support analysis based on models developed by data mining. The architecture and implementation of the model for real-time process control based on the integration of data mining and OLAP is presented in detail. The model is validated by comparing results obtained from the integrated system, OLAP-only and expert opinion. The system is validated using actual process data and the results of this verification are presented. A discussion of the results of the validation of the integrated system and some limitations of this research with discussion on possible future research directions is provided.
21
Cakici, Avni. "Self-tuning PI control of industrial wastewater treatment process and simultaneous on-line estimations of oxygen uptake and transfer rates." Thesis, University of Exeter, 1992. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.317288.
Full text
22
Rocha, Andre Luiz Sena da. "Controle on-line por atributos para o n?mero de n?o-conformidades no item inspecionado com base em uma sequ?ncia de inspe??o." Universidade Federal do Rio Grande do Norte, 2010. http://repositorio.ufrn.br:8080/jspui/handle/123456789/15019.
Full textAbstract:
Made available in DSpace on 2014-12-17T14:53:03Z (GMT). No. of bitstreams: 1
AndreLSR_DISSERT.pdf: 1376512 bytes, checksum: 77587a4d46a8d965b17b7df02200c5a6 (MD5)
Previous issue date: 2010-12-29Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior
This paper proposes a procedure to control on-line processes for attributes, using an Shewhart control chart with two control limits (warning limit and control limit) and will be based on a sequence of inspection (h). The inspection procedure is based on Taguchi et al. (1989), in which to inspect the item, if the number of non-conformities is higher than an upper control limit, the process needs to be stopped and some adjustment is required; and, if the last inspection h, from all items inspected present a number of non-conformities between the control limit and warning limit. The items inspected will suffer destructive inspection, being discarded after inspection. Properties of an ergodic Markov chain are used to get the expression of average cost per item and the aim was the determination of four optimized parameters: the sampling interval of the inspections (m); the constant W to draw the warning limit (W); the constant C to draw the control limit (C), where W ? C, and the length of sequence of inspections (h). Numerical examples illustrate the proposed procedure
Este trabalho prop?e um procedimento de controle on-line de processos por atributos, utilizando um gr?fico de Shewhart com dois Limites de Controle (Limite de Advert?ncia e Limite de Controle) e ser? baseado numa sequ?ncia de inspe??o (h). O procedimento de inspe??o ? baseado em Taguchi et al. (1989). Assim, ao inspecionar o item, o processo ? parado para ajuste se o n?mero de n?o conformidades for superior ao Limite de Controle; como tamb?m se nas ?ltimas h inspe??es, todos os h itens inspecionados apresentarem um n?mero de n?o-conformidades entre os Limites de Advert?ncia e de Controle. Utilizando-se de propriedades de uma Cadeia de Markov Erg?dica, foi obtida uma express?o do custo m?dio do sistema de controle, que pode ser minimizada por quatro par?metros: Intervalo entre inspe??es (m), Limite de Advert?ncia (W), Limite de controle (C, em que W ≤ C) e o tamanho da sequ?ncia de inspe??o (h). Um exemplo num?rico ilustra o procedimento proposto
23
Rodrigues, Renata Mendon?a. "Controle on-line para o n?mero de n?o-conformidades em um ?tem inspecionado." Universidade Federal do Rio Grande do Norte, 2009. http://repositorio.ufrn.br:8080/jspui/handle/123456789/18628.
Full textAbstract:
Made available in DSpace on 2015-03-03T15:22:32Z (GMT). No. of bitstreams: 1
RenataMR.pdf: 665159 bytes, checksum: 6d95fdc9fbeffbdaabe5793488839df3 (MD5)
Previous issue date: 2009-03-02Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior
The on-line processes control for attributes consists of inspecting a single item at every m produced ones. If the examined item is conforming, the production continues; otherwise, the process stops for adjustment. However, in many practical situations, the interest consist of monitoring the number of non-conformities among the examined items. In this case, if the number of non-conformities is higher than an upper control limit, the process needs to be stopped and some adjustment is required. The contribution of this paper is to propose a control system for the number of nonconforming of the inspected item. Employing properties of an ergodic Markov chain, an expression for the expected cost per item of the control system was obtained and it will be minimized by two parameters: the sampling interval and the upper limit control of the non-conformities of the examined item. Numerical examples illustrate the proposed procedure
O procedimento usual de controle on-line de processo por atributos consiste em inspecionar um item a cada m itens produzidos. Se o item examinado for conforme, a produ??o continua; caso contr?rio p?ra-se o processo. No entanto, em muitas situa??es pr?ticas, nem sempre existe interesse em classificar o item como defeituoso ou n?o defeituoso, mas sim monitorar o n?mero de n?o-conformidades no item inspecionado. Neste caso, se o n?mero de n?o-conformidades for superior a um limite de controle, p?ra-se o processo para o ajuste. A contribui??o deste trabalho est? em propor um sistema de controle on-line baseado no n?mero de n?o-conformidades do item inspecionado. Atrav?s das propriedades de uma cadeia de Markov erg?dica, obteve-se uma express?o anal?tica do custo m?dio por item produzido do sistema de controle on-line que pode ser minimizada por dois par?metros: o intervalo entre inspe??es e o limite superior de controle para o n?mero de n?o-conformidades no item inspecionado. Um exemplo num?rico ilustra o procedimento proposto
24
Kohlert, Michael [Verfasser], Andreas [Akademischer Betreuer] König, and Tamara [Akademischer Betreuer] Chistyakova. "Multi-Sensory Data Analysis and On-Line Evaluation for Advanced Process Control and Yield Optimization in Polymer Film Industry / Michael Kohlert. Betreuer: Andreas König ; Tamara Chistyakova." Kaiserslautern : Technische Universität Kaiserslautern, 2015. http://d-nb.info/1077007140/34.
Full text
25
Beaupré, Mathieu. "Characterization of on-line sensors for water quality monitoring and process control = : Caractérisation des capteurs en ligne dans le domaine de la qualité de l'eau et du contrôle de prodécé." Master's thesis, Université Laval, 2010. http://hdl.handle.net/20.500.11794/22104.
Full textAbstract:
L'arrivée des nouvelles technologies et des ordinateurs en plus des problématiques grandissantes reliées à la qualité de l'eau ont fait que le monitorage dans le domaine de la qualité de l'eau a littéralement explosé au cours des dernières décennies. Le monitorage consiste à effectuer le suivi d'un système et de faire la conservation des données récoltées. Il est habituellement réalisé à l'aide de capteurs. Ces derniers peuvent générer une importante quantité de données. Dans certains cas, la fréquence d'échantillonnage peut atteindre plusieurs mesures à la seconde. Nonobstant la quantité de données recueillies, il faut en connaître la qualité afin de les utiliser. Quels sont les outils disponibles pour caractériser les capteurs? Quelles informations fournissent-ils? Est-ce que ces dernières sont suffisantes? Certains protocoles pour évaluer la performance des capteurs sont disponibles. Le protocole ISO 15839:2003, Qualité de l'eau - Matériel d'analyse/capteurs directs pour l'eau - Spécifications et essais de performance, est le plus complet à ce jour. Il permet d'évaluer la performance des capteurs sous des conditions de laboratoire. Néanmoins, est-ce que de soumettre les capteurs à ce protocole permet aux usagers de choisir le capteur le plus approprié? Afin de répondre à cette question, la première étape de l'étude consistait à passer en revue les capteurs selon les technologies disponibles sur le marché. L'accent fut mis sur les substances suivantes : nitrate, nitrite, ammonium, phosphore, oxygène dissout et la turbidité. La revue a révélé que les fabricants ne fournissent pas ou très peu de spécifications découlant du protocole ISO 15839:2003. La deuxième étape fut de procéder à une revue critique du protocole ISO 15839:2003 afin de faire ressortir les raisons pour lesquelles les fabricants n'en fournissent pas les spécifications et de vérifier si ces dernières sont adéquates pour choisir le bon capteur pour une application précise. Les spécifications résultant de l'application du protocole ISO 15839:2003 et la façon dont elles sont présentées ne permettent pas aux usagers de choisir le capteur le plus approprié pour leur application. Le protocole devrait davantage être orienté sur les besoins des usagers plutôt que sur ceux des fabricants. À l'instar des profiles d'exactitudes (Hubert et al., 2004), le protocole ISO 15839:2003 devrait chercher davantage à présenter lesrésultats sous forme graphique. Le protocole ISO 15839:2003 procure les spécifications des capteurs sous des conditions de laboratoire, mais il contient aussi une section traitant des spécifications sous des conditions de terrain (field conditions). Toutefois, les résultats sont subordonnés au temps et à l'emplacement où les tests sont effectués, ce qui rend impossible la comparaison des résultats provenant de différents capteurs. La revue de littérature a démontré la nécessité de développer un protocole de tests se faisant sous des conditions de terrain reproduites de façon standard. Deux interférences rencontrées sur le terrain furent reproduites en laboratoire : la turbidité ainsi que les bulles d'air retrouvées dans les bassins d'aération des usines de traitement des eaux usées. Les résultats ont démontré que ces deux interférences amplifient le bruit de mesure et génèrent un biais. En quantifiant le bruit de mesure et le biais les capteurs peuvent être comparés, ce qui permet aux usagers de sélectionner le capteur le plus approprié.
26
Schaefer, Cédric. "A Process Analytical Technology (PAT) approach involving near infrared spectroscopy to control the manufacturing of an active pharmaceutical ingredient : development, validation and implementation." Thesis, Aix-Marseille, 2013. http://www.theses.fr/2013AIXM4375.
Full textAbstract:
Les entreprises pharmaceutiques ont progressivement adopté le concept de Process Analytical Technology (PAT) afin de contrôler et d'assurer en temps réel la qualité des produits pharmaceutiques au cours de leur production. Le PAT et un composant central du concept plus général de Quality-by-Design (QbD) promu par les agence régulatrices et visant à construire la qualité des produits via une approche scientifique et la gestion des risques.Une méthode basée sur la spectroscopie proche infrarouge (PIR) a été développée comme un outil du PAT pour contrôler en ligne la cristallisation d'un principe actif pharmaceutique. Au cours du procédé les teneurs en principe actif et en solvant résiduel doivent être déterminées avec précision afin d'atteindre un point d'ensemencement prédéfini. Une méthodologie basée sur les principes du QbD a guidé le développement et la validation de la méthode tout en assurant l'adéquation avec son utilisation prévue. Des modèles basés sur les moindres carrés partiels ont été construits à l'aide d'outils chimiométriques afin de quantifier les 2 analytes d'intérêt. La méthode a été totalement validée conformément aux requis officiels en utilisant les profils d'exactitude. Un suivi du procédé en temps réel a permis de prouver que la méthode correspond à son usage prévu.L'implémentation de cette méthode comme à l'échelle industrielle au lancement de ce nouveau procédé permettra le contrôle automatique de l'étape de cristallisation dans le but d'assurer un niveau de qualité prédéfini de l'API. D'autres avantages sont attendus incluant la réduction du temps du procédé, la suppression d'un échantillonnage difficile et d'analyses hors ligne fastidieusesPharmaceutical companies are progressively adopting and introducing the Process Analytical Technology (PAT) concept to control and ensure in real-time product quality in development and manufacturing. PAT is a key component of the Quality-by-Design (QbD) framework promoted by the regulatory authorities, aiming the building of product quality based on both a strong scientific background and a quality risk management approach.An analytical method based on near infrared (NIR) spectroscopy was developed as a PAT tool to control on-line an API (active pharmaceutical ingredient) crystallization. During this process the API and residual solvent contents need to be precisely determined to reach a predefined seeding point. An original methodology based on the QbD principles was applied to conduct the development and validation of the NIR method and to ensure that it is fitted for its intended use. Partial least squares (PLS) models were developed and optimized through chemometrics tools in order to quantify the 2 analytes of interest. The method was fully validated according to the official requirements using the accuracy profile approach. Besides, a real-time process monitoring was added to the validation phase to prove and document that the method is fitted for purpose.Implementation of this method as an in-process control at industrial plant from the launch of this new pharmaceutical process will enable automatic control of the crystallization step in order to ensure a predefined quality level of the API. Other valuable benefits are expected such as reduction of the process time, and suppression of a difficult sampling and tedious off-line analyzes
27
Rice, Matthew. "New Techniques for Continuous Chemical Analysis in the Pulp&Paper Industry." Doctoral thesis, Stockholm : Tekniska högsk, 2001. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-3131.
Full text
28
Silva, Paulo Henrique Ferreira da. "Introdução ao controle estatístico de processo on-line." Universidade Federal de São Carlos, 2011. https://repositorio.ufscar.br/handle/ufscar/4553.
Full textAbstract:
Made available in DSpace on 2016-06-02T20:06:05Z (GMT). No. of bitstreams: 1
4182.pdf: 2255930 bytes, checksum: 7da65af2bc4550af129f4ceaea0a8eb1 (MD5)
Previous issue date: 2011-04-07Financiadora de Estudos e Projetos
Neste trabalho são apresentadas algumas ferramentas do Controle Estatístico de Processos (CEP), que podem ser usadas no monitoramento de sistemas produtivos ao longo do tempo, bem como a sua aplicação em conjuntos de dados artificiais, que fazem parte de contextos reais. O estudo sobre essas ferramentas estatísticas está distribuído de tal maneira que são abordados durante o trabalho: a teoria das ferramentas estudadas, os diferentes contextos em que podem ser aplicadas e a sua implementação on-line, utilizando recursos computacionais de softwares livres. A aplicação no sistema on-line é realizada de tal modo que viabiliza a praticidade e eficácia na geração de gráficos para o CEP e de índices que refletem a capacidade do processo. É apresentada também uma sequência de passos para o uso do sistema de CEP on-line aqui proposto, tanto para situações em que se observa apenas uma medida da qualidade do processo (caso univariado), quanto para situações em que são observadas várias medidas (caso multivariado).
29
Duarte, Elis Regina. "Avaliação e desenvolvimento de algoritmos de controle aplicado a um processo extrativo de fermentação alcoolica continua." [s.n.], 2007. http://repositorio.unicamp.br/jspui/handle/REPOSIP/267016.
Full textAbstract:
Orientadores: Rubens Maciel Filho, Laercio EnderTese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Quimica
Made available in DSpace on 2018-08-08T18:03:43Z (GMT). No. of bitstreams: 1 Duarte_ElisRegina_D.pdf: 5707651 bytes, checksum: d1348f1129ff1981af5a94400a0f9040 (MD5) Previous issue date: 2007
Resumo:0 objetivo deste trabalho foi desenvolver e avaliar diferentes algoritmos de controle para o processo extrativo de fermentação alcoólica contínua. Para isto foram comparados controladores do tipo preditivo e adaptativo. Para o controle preditivo, foi avaliado o Controle por Matriz Dinâmica (DMC) e foi desenvolvido um algoritmo de controle preditivo baseado em modelo usando redes neurais artificiais (MPC Neural) com aprendizagem em tempo real das redes. Para o controle adaptativo, foi proposto o aperfeiçoamento do algoritmo de controle CONDEG (Controle Neural Direto Baseado no Erro Global) Modificado, desenvolvido por Duarte (2004), O algoritmo está baseado em redes neurais artificiais, com aprendizagem em tempo real, de acordo com as alterações que ocorrem no processo. Os parâmetros de penalização das ações de controle, que são parâmetros de projeto do controlador, foram ajustados ao longo do tempo através da aplicação de um algoritmo do Filtro de Kalman. Para o procedimento de investigação foi utilizada a simulação computacional para o qual todos os algoritmos de controle estudados foram implementados em linguagem de programação Fortran 90 e aplicados a um processo extrativo de fermentação alcoólica contínua para produção de etanol desenvolvido por Silva (1999). O modelo matemático utilizado foi desenvolvido por Costa et al(2001). As simulações em malha fechada realizadas utilizando os algoritmos propostos mostraram melhores resultados para os algoritmos de controle usando redes com aprendizagem ao longo do tempo e que o algoritmo de controle CONDEG Modificado usando filtro de Kalman com fator de velocidade associado foi eficiente e robusto, pois apresentou bons resultados em problemas dos tipos servos e regulador.
Abstract: The objective of the present work is to develop and to evaluate the performance of predictive and adaptive controllers, applied to an extractive fermentative process. As predictive controllers the Dynamical Matrix Control (DMC) and a model predictive control based on artificial neural networks with on-line learning were considered. The adaptive controller is an improvement of the Modified Condeg strategy control (Direct Neural Control based on Global Error), developed by Duarte (2004). The strategy is based on artificial neural networks, with on-line learning, according to modifications that occur in the process. The control actions penalization parameters, that are in fact controller design parameters, are on-line adjusted through an algorithm based on Kalman filter. The performance evaluation was carried out through computer simulation with all algorithms implemented in Fortran 90,.As a case study, an extractive fermentation alcoholic process developed by Silva (1999) was taken into account with the mathematical model developed by Costa et. al (2001). The results obtained from closed-loop simulations using the proposed algorithms showed better results for the neural networks with on-line learning. The Modified Condeg wsth Kalman Filter plus velocity factor is efficient and robust for servo and regulatory applications.
Doutorado
Desenvolvimento de Processos Químicos
Doutor em Engenharia Química
30
Björk, Anders. "Chemometric and signal processing methods for real time monitoring and modeling : applications in the pulp and paper industry." Doctoral thesis, KTH, Kemi, 2007. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-4383.
Full textAbstract:
In the production of paper, the quality of the pulp is an important factor both for the productivity and for the final quality. Reliable real-time measurements of pulp quality are therefore needed. One way is to use acoustic or vibration sensors that give information-rich signals and place the sensors at suitable locations in a pulp production line. However, these sensors are not selective for the pulp properties of interest. Therefore, advanced signal processing and multivariate calibration are essential tools. The current work has been focused on the development of calibration routes for extraction of information from acoustic sensors and on signal processing algorithms for enhancing the information-selectivity for a specific pulp property or class of properties. Multivariate analysis methods like Principal Components Analysis (PCA), Partial Least Squares (PLS) and Orthogonal Signal Correction (OSC) have been used for visualization and calibration. Signal processing methods like Fast Fourier Transform (FFT), Fast Wavelet Transform (FWT) and Continuous Wavelet Transform (CWT) have been used in the development of novel signal processing algorithms for extraction of information from vibrationacoustic sensors. It is shown that use of OSC combined with PLS for prediction of Canadian Standard Freeness (CSF) using FFT-spectra produced from vibration data on a Thermo Mechanical Pulping (TMP) process gives lower prediction errors and a more parsimonious model than PLS alone. The combination of FFT and PLS was also used for monitoring of beating of kraft pulp and for screen monitoring. When using regular FFT-spectra on process acoustic data the obtained information tend to overlap. To circumvent this two new signal processing methods were developed: Wavelet Transform Multi Resolution Spectra (WT-MRS) and Continuous Wavelet Transform Fibre Length Extraction (CWT-FLE). Applying WT-MRS gave PLS-models that were more parsimonious with lower prediction error for CSF than using regular FFT-Spectra. For a Medium Consistency (MC) pulp stream WT-MRS gave predictions errors comparable to the reference methods for CSF and Brightness. The CWT-FLE method was validated against a commercial fibre length analyzer and good agreement was obtained. The CWT-FLE-curves could therefore be used instead of other fibre distribution curves for process control. Further, the CWT-FLE curves were used for PLS modelling of tensile strength and optical parameters with good results. In addition to the mentioned results a comprehensive overview of technologies used with acoustic sensors and related applications has been performed.Vid framställning av pappersprodukter är kvaliteten på massan en viktig faktor för produktiviteten och kvalitén på slutresultatet. Det är därför viktigt att ha tillgång till tillförlitliga mätningar av massakvalitet i realtid. En möjlighet är att använda akustik- eller vibrationssensorer i lämpliga positioner vid enhetsoperationer i massaprocessen. Selektiviteten hos dessa mätningar är emellertid relativt låg i synnerhet om mätningarna är passiva. Därför krävs avancerad signalbehandling och multivariat kalibrering. Det nu presenterade arbetet har varit fokuserat på kalibreringsmetoder för extraktion av information ur akustiska mätningar samt på algoritmer för signalbehandling som kan ge förbättrad informationsselektivitet. Multivariata metoder som Principal Component Analysis (PCA), Partial Least Squares (PLS) and Orthogonal Signal Correction (OSC) har använts för visualisering och kalibrering. Signalbehandlingsmetoderna Fast Fourier Transform (FFT), Fast Wavelet Transform (FWT) och Continuous Wavelet Transform (CWT) har använts i utvecklingen av nydanande metoder för signalbehandling anpassade till att extrahera information ur signaler från vibrations/akustiska sensorer. En kombination av OSC och PLS applicerade på FFT-spektra från raffineringen i en Termo Mechnaical Pulping (TMP) process ger lägre prediktionsfel för Canadian Standard Freeness (CSF) än enbart PLS. Kombinationen av FFT och PLS har vidare använts för monitorering av malning av sulfatmassa och monitorering av silning. Ordinära FFT-spektra av t.ex. vibrationssignaler är delvis överlappande. För att komma runt detta har två signalbehandlingsmetoder utvecklats, Wavelet Transform Multi Resolution Spectra (WT-MRS) baserat på kombinationen av FWT och FFT samt Continuous Wavelet Transform Fibre Length Extraction (CWT-FLE) baserat på CWT. Tillämpning av WT-MRS gav enklare PLS-modeller med lägre prediktionsfel för CSF jämfört med att använda normala FFT-spektra. I en annan tillämpning på en massaström med relativt hög koncentration (Medium Consistency, MC) kunde prediktioner för CSF samt ljushet erhållas med prediktionsfel jämförbart med referensmetodernas fel. Metoden CWT-FLE validerades mot en kommersiell fiberlängdsmätare med god överensstämmelse. CWT-FLE-kurvorna skulle därför kunna användas i stället för andra fiberdistributionskurvor för processtyrning. Vidare användes CWT-FLE kurvor för PLS modellering av dragstyrka samt optiska egenskaper med goda resultat. Utöver de nämnda resultaten har en omfattande litteratursammanställning gjorts över området och relaterade applikationer.
QC 20100629
31
Cunha, Andre Pitasse da. "Automação e controle "on line" de uma coluna de destilação em batelada." [s.n.], 1996. http://repositorio.unicamp.br/jspui/handle/REPOSIP/267305.
Full textAbstract:
Orientadores: João Alexandre Ferreira da Rocha Pereira, Ana Maria Frattini FiletiDissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Quimica
Made available in DSpace on 2018-07-21T03:47:55Z (GMT). No. of bitstreams: 1 Cunha_AndrePitasseda_M.pdf: 2972168 bytes, checksum: 74e7ec31a3e97ab0d9d409c9d1de8d2a (MD5) Previous issue date: 1996
Resumo: No presente trabalho, foi desenvolvido um sistema de controle adaptativo digital baseado em planejamento da razão de refluxo para a destilação em batelada do binário n.hexano-n.heptano, objetivando o controle da composição molar de topo do componente mais volátil dentro de uma especificação desejada. Para tal projeto, utilizou-se um programa de simulação da coluna de destilação em estudo com operação contínua para caracterização das, etapas da destilação em batelada, através da variação da composição molar de hexano no refervedor, simulando a escassez ocorrida durante a operação em batelada. E pelo método da curva de reação de Cohen e Coon, levantaram-se os perfis dos parâmetros do controlador. O algoritmo de controle foi testado em simulação computacional (off-line) e corrida experimental (on-line) da destilação em batelada. Os ajustes encontrados foram excelentes
Abstract: In this work, a digital adaptative control system was developed based on a reflux rate policy to the batch distillation of the system n-hexane and n-heptane binary, aiming at the control of top molar composition of the more volatile component within a desired specification. For this project, a simulation program of a distillation column under continuous operation was used in order to characterize the steps of the batch distillation, through the variation of the hexane molar composition at the reboiler, simulating the scarcity occured during batch operation. By using the Process Reaction Curve Method of Cohen and Coon, the instantaneous values of the controller parameters were obtained. The control algorithm was tested through computational simulation and in a pilot batch column. The agreement was found to be excellent
Mestrado
Mestre em Engenharia Química
32
Oisiovici, Ronia Marques. "Operação e controle por computador "on-line" de uma coluna de destilação em batelada." [s.n.], 1998. http://repositorio.unicamp.br/jspui/handle/REPOSIP/266409.
Full textAbstract:
Orientadores: Sandra Lucia da Cruz, João Alexandre F. R. PereiraDissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Quimica
Made available in DSpace on 2018-07-23T13:37:35Z (GMT). No. of bitstreams: 1 Oisiovici_RoniaMarques_M.pdf: 4409740 bytes, checksum: 4164f154c242c905c105f8aac5858d50 (MD5) Previous issue date: 1998
Resumo: Os benefícios a serem ganhos com a automação de processos em batelada são muitos: aumento de produção, melhoria de qualidade, redução de custos e maior segurança operacional. A destilação em batelada, que está sendo largamente empregada na indústria, oferece grande flexibilidade de operação, porém o seu caráter não-linear e não-estacionário faz com que o controle e a automação se tomem um desafio. Os trabalhos experimentais sobre destilação em batelada são muito poucos, principalmente na área de controle. Além disso, os sistemas estudados apresentam comportamento ideal, não se encontrando nada na literatura a respeito de sistemas não-ideais. Considerando a carência de trabalhos práticos, o interesse pela automação de processos e a crescente importância da destilação em batelada, o presente trabalho estudou experimentalmente a operação de uma coluna de destilação em batelada em escala piloto controlada automaticamente por um microcomputador conectado "on-line" ao processo. A interface entre o micro e o processo foi feita através de uma placa conversora Analógica-Digital/Digital-Analógica (AD/DA). Foram desenvolvidas estratégias de controle para manter constante a composição do destilado (variável controlada) variando-se a razão de refluxo (variável manipulada). A primeira etapa do trabalho consistiu na elaboração de um programa simulador da coluna, com o qual foram testadas diferentes estratégias de controle. Três estratégias que se mostraram eficientes a nível de simulação foram, então, implementadas na prática. São elas: controle adaptativo auto-ajustável, controle por antecipação baseado no balanço de massa e controle proporcional baseado na simulação da coluna. Para implementar experimentalmente os algoritmos de controle, foi necessário considerar alguns aspectos não presentes a nível de simulação, tais como inferência da composição e presença de ruídos nos sinais.Devido ao alto custo e complexidade dos medidores de composição, desenvolveu-se um método para inferência da composição de sistemas binários não-ideais a partir da medida de temperatura. Um termopar foi conectado na base e outro no topo da coluna. Para diminuir o nível de ruído da temperatura lida pelo computador, foram feitos testes com alguns filtros digitais. As sub-rotinas de filtragem digital e inferência da composição foram inseridas no programa de aquisição de dados e controle. A cada período de amostragem, as saídas analógicas dos sensores de temperatura eram amplificadas, convertidas em sinais digitais, lidas e filtradas pelo computador. As temperaturas lidas eram utilizadas para inferência da composição da mistura no refervedor e do vapor no topo da coluna. A partir dos dados atuais e, a depender da estratégia de controle escolhida, dos dados passados, calculava-se a razão de refluxo necessária para manter a pureza do destilado no valor especificado.A razão de refluxo era, então, implementada através de uma válvula magnética controlada pelo computador. Os testes foram feitos com o sistema não-ideal etanol/água e com o sistema ideal etanol/l-propanol. Os três controladores apresentaram bom desempenho. Comparando-se os resultados obtidos com as três estratégias de controle, pode-se concluir que, para uma mesma precisão de controle, a diferença entre os tempos de batelada não é significativa. Porém, o perfil de razão de refluxo do controlador por antecipação é mais suave do que os perfis obtidos com os controladores adaptativo e proporcional. A influência do nível de ruído das medidas de temperatura sobre a inferência da composição depende das propriedades termodinâmicas da mistura a ser destilada. Para o sistema etanol/água na faixa X IND. EtOH > 0,60 , foram observadas grandes oscilações nos valores de composição pois, nesse caso, a precisão de + ou - 1°C não é suficiente para uma boa inferência.No entanto, para o binário etanol/l-propanol, essa precisão mostrou ser adequada para toda faixa de composição (O < ou = X IND. EtOHS < ou = 1,0)
Abstract: The benefits to be gained from automation of batch processes are many: increased production, improved product quality, lower costs and greater safety. Batch distillation, which has been widely used in industry, provides operating flexibility but its non-linear and unsteady behaviour makes control and automation a challenge. Experimental works on batch distillation are rare, mainly in control area, and they are limited to ideal systems. Works with non-ideal systems have not been found in literature. Considering the lack of practical works, the interest in process automation and the growing importance of batch distillation, the present work studied experimentally the operation of a pilot-scale batch distillation column controlled by a microcomputer connected on-line to the process. The interface between the computer and the process was achieved by an AD/DA converter. Control strategies were developed to keep the distillate purity (controlled variable) constant by varying the reflux ratio (manipulated variable). The first step was the development of a batch distillation column simulator to test different control strategies. Simulation results of three controllers were good, so they were implemented experimentally: a self-tuning regulator (STR), a feed forward controller (FC) based on mass balance and a proportional controller (PC) based on simulation results. Some issues which were not taken into account in the column simulation, such as inference of composition and presence of noise, had then to be considered. Because of the high cost and complexity of on-line analysers, a methodology was developed to infer compositions of non-ideal binary systems from temperature measurements. The top column and still temperatures were measured with thermocouples. In order to reduce the noise leveI, different digital filters were implemented and tested. At each sampling period, the computer sent a digital signal to select the top or bottom thermo couple channel. The analogue output signals of the temperature sensors were amplified, converted to digital signals, read and filtered by the computer. The temperature measurements were used to infer the top vapour composition and the still composition. The current data and, depending on the chosen control strategy, the past data were used to calculate the reflux ratio to keep the distillate purity at the set-point. The reflux ratio was then implemented with a magnetic valve controlled by the computer. Tests were carried out with the non-ideal system ethanol/water and with the ideal system ethanol/l-propanol. The three controllers have shown good performance. For the same control accuracy, the difference in batch time was negligible.The main difference was found to be in the reflux ratio profile. The FC's reflux ratio profiles were smoother than the ones obtained with the STR and with the Pc. The influence of the noise of temperature measurements on the inferred compositions depends on the thermodynamic properties of each system. For the ethanol/water system, a higher leveIof oscillationswas present when the top vapour composition was above 0.6 because the available precision of + or ¿ 1ºC was not good enough to infer accurate composition values. However, for the ethanol/l-propanol system, this precision has shown to be suitable for the full composition range
Mestrado
Sistemas de Processos Quimicos e Informatica
Mestre em Engenharia Química
33
Pierret, Vanusa Beatriz Hackenhaar. "DESENVOLVIMENTO DE UM APLICATIVO DO CONTROLE ESTATÍSTICO DO PROCESSO ON-LINE POR MEIO DE GRÁFICOS DE CONTROLE." Universidade Federal de Santa Maria, 2004. http://repositorio.ufsm.br/handle/1/8033.
Full textAbstract:
Coordenação de Aperfeiçoamento de Pessoal de Nível SuperiorThe economy globalization forced the companies to invest in other forms of management and in technologies that intensify its competitiveness on the market and provide an increasing in the quality and in the cost/benefit of their products and services. In this context, the purpose of this research consists of creating an interactive site about Statistical Process Control (SPC), that could help companies to have theoretical-practical notions of the operation of this tool and of its application. In order to get that, the tools of Statistical Process Control will be used, helping companies to forecast the output trials and, at the same time, increase the quality of its products and services, avoiding scraps and redo. In this way, the concepts and the software used in the analysis, construction and implementation of the interactive site are described. Subsequently, the created tool has been tested and evaluated, by mean a cases. Results obtained revealed that it is efficient, because the graphics generated shown the user a real and concrete vision based on the supplied information for the CEP on-line program. Finally, based on the results, a deeper study regarding some limitations presented by the tool must continue, like the size of the sample and review calculus of limits, which are not performed by the program.
Com a crescente globalização da economia e abertura de novos mercados, as empresas estão sendo impulsionadas a investir em outras formas de gestão e em tecnologias que intensifiquem a sua competitividade no mercado e proporcionem um aumento na qualidade e no custo/benefício de seus produtos e serviços. Neste contexto, o objetivo deste trabalho consiste em criar um site1 interativo sobre o Controle Estatístico do Processo (CEP), o qual possa contribuir para que as empresas tenham noções teórico-práticas do funcionamento dessa ferramenta e de sua aplicação. Para tanto, serão utilizadas as ferramentas do Controle Estatístico do Processo (CEP), visando auxiliar as empresas a tornar previsíveis as saídas de seus processos de produção e, ao mesmo tempo, aumentar a qualidade de seus produtos e serviços, evitando o refugo e o retrabalho. Dessa forma, descrevem-se os conceitos e os softwares2 utilizados, na análise, construção e implementação do site interativo. Posteriormente, o aplicativo criado foi testado e avaliado, utilizando-se para isso amostras de estudos de casos. Os resultados obtidos revelaram que é eficiente, uma vez que os gráficos gerados permitem ao usuário uma visão real e concreta baseada nas informações fornecidas ao aplicativo CEP on-line . Finalmente, com base nos resultados, deve-se proceder a um estudo mais aprofundado com relação a algumas limitações que o aplicativo apresenta, tais como o tamanho da amostra, pois se restringe a amostras somente de tamanho constante, e os cálculos dos limites revisados, os quais não são efetuados pelo aplicativo.
34
Silva, Lenilson Pereira da. "Ajuste preventivo versus ajuste corretivo no controle on-line de processo do n?mero de n?o-conformidades num item inspecionado." Universidade Federal do Rio Grande do Norte, 2010. http://repositorio.ufrn.br:8080/jspui/handle/123456789/18637.
Full textAbstract:
Made available in DSpace on 2015-03-03T15:28:32Z (GMT). No. of bitstreams: 1
LenilsonPS_DISSERT.pdf: 421189 bytes, checksum: 22f993fe2b8033189672f2dec900232c (MD5)
Previous issue date: 2010-06-15In production lines, the entire process is bound to unexpected happenings which may cost losing the production quality. Thus, it means losses to the manufacturer. Identify such causes and remove them is the task of the processing management. The on-line control system consists of periodic inspection of every month produced item. Once any of those items is quali ed as not t, it is admitted that a change in the fraction of the items occurred, and then the process is stopped for adjustments. This work is an extension of Quinino & Ho (2010) and has as objective main to make the monitoramento in a process through the control on-line of quality for the number of non-conformities about the inspected item. The strategy of decision to verify if the process is under control, is directly associated to the limits of the graphic control of non-conformities of the process. A policy of preventive adjustments is incorporated in order to enlarge the conforming fraction of the process. With the help of the R software, a sensibility analysis of the proposed model is done showing in which situations it is most interesting to execute the preventive adjustment
Numa linha de produ??o, todo processo est? sujeito a ocorr?ncias de causas que venham a aparecer causando perdas na qualidade do produto e consequentemente preju?zos ao fabricante. Identificar essas causas e remov?-las ? tarefa do respons?vel pelo monitoramento do processo. Este trabalho tem como objetivo principal fazer o monitoramento num processo atrav?s do controle on-line de qualidade por atributos, considerando o n?mero de n?o-conformidades no item inspecionado. A estrat?gia de decis?o para verificar se o processo est? sob controle, est? diretamente associado aos limites de controle do gr?fico de n?o-conformidades do processo. Uma pol?tica de ajustes preventivos ? incorporada com o objetivo de aumentar a fra??o conforme do processo. Com o aux?lio do software R, ? feita uma an?lise de sensibilidade do modelo proposto mostrando em que situa??es ? mais interessante fazer o ajuste preventivo
35
Ávila, Thiago Carvalho de 1985. "Espectroscopia Raman e quimiometria como ferramentas no monitoramento on-line do processo fermentativo da glicose pela Saccharomyces cerevisiae." [s.n.], 2013. http://repositorio.unicamp.br/jspui/handle/REPOSIP/249299.
Full textAbstract:
Orientador: Ronei Jesus PoppiDissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Química
Made available in DSpace on 2018-08-22T08:18:21Z (GMT). No. of bitstreams: 1 Avila_ThiagoCarvalhode_M.pdf: 7831860 bytes, checksum: 010f2295e00f097a9ecfaf3f498a7069 (MD5) Previous issue date: 2013
Resumo: Este trabalho visou o uso de Espectroscopia Raman e de Quimiometria para monitoramento e controle da fermentação de glicose por Saccharomyces cerevisiae. Na primeira etapa, foi utilizada calibração multivariada baseada no método dos Mínimos Quadrados Parciais (PLS) para quantificação de glicose, etanol, glicerol, ácido acético e células. Os modelos foram desenvolvidos baseados nos valores de concentração obtidos pelos métodos de referência, cromatografia líquida de alta eficiência ¿ HPLC e espectrofotometria UV/Vis. Tanto na etapa de calibração quanto na de validação, a otimização foi realizada com eliminação de amostras anômalas, baseada nos valores de leverage, resíduos e escores. Na segunda etapa, cartas de controle multivariadas foram usadas para identificação de falhas em bateladas durante o processo de fermentação. Foram construídos modelos MPCA (Análise de Componentes Principais Multimodo) a partir de bateladas NOC (Condições Normais de Operação). As cartas de controle multivariadas foram aplicadas em dois modos de desdobramento dos dados obtidos durante o monitoramento, um preservando a direção das bateladas e outro a direção do tempo. As falhas estudadas foram temperatura, mudança no substrato e contaminação do sistema. No modo de desdobramento por bateladas, a carta de controle Q foi eficiente para detecção das falhas estudas, fato comprovado pela classificação correta de três bateladas NOC como dentro de controle. No entanto, a carta de controle T2 não foi capaz de identificar as falhas estudadas corretamente como fora de controle. O modo de desdobramento pelo tempo também apresentou classificações corretas das falhas estudadas
Abstract: This work aims the use of Raman Spectroscopy and Chemometrics in the monitoring and control in the fermentation of the glucose by Saccharomyces cerevisiae. In the first step, it was applied the multivariate calibration based on Partial Least Squares (PLS) for the quantification of glucose, ethanol, glycerol, acetic acid and cells. The developed of calibration models was performed against the concentration values obtained by the reference methods, High Performance Liquid Chromatography and UV/Vis spectrophotometer. The optimization of the calibration and validation steps, the elimination of outliers was performed based on the values of leverage, residues and scores. In the second step, multivariate control charts were used for identification of batch-fault during the fermentation process. Multi-way Principal Component Analysis (MPCA) models were developed from batch NOC (Normal Operation Conditions). The multivariate control charts were based on two modes of unfolding the multi-way data, obtained during monitoring, one preserving the direction of the batch and another the direction of time. The fault studied were temperature, changes in the substrate and contamination of the system. In unfolding batch mode, the chart Q was effective for detection of the faults studied, proven by the correctly classification of 3 NOC batches as in control. However, the chart T2 failed to identify faults studied. The unfolding in time mode, also presented correct classifications of the faults studied
Mestrado
Quimica Analitica
Mestre em Química
36
Ñahuincopa, Arango Julio. "Sistematización del proceso de cálculo de presupuestos, rendimiento de maquinarias y programación lineal para movimiento de tierras con la implementación de una aplicación on line." Bachelor's thesis, Universidad Continental, 2016. http://repositorio.continental.edu.pe/handle/continental/3034.
Full textAbstract:
El tema que se aborda en la presente tesis es la sistematización de los procesos de cálculo de presupuesto, rendimiento de maquinarias y programación lineal mediante la implementación de una aplicación on line. De acuerdo a los objetivos planteados se desarrolla procedimientos que permiten implementar la aplicación on line. Para lograr tal objetivo, es necesario la intervención de una metodología de la tecnología de la información.Tesis
37
Creti, Christian. "Fermentation méthanique et désulfuration de gaz par voie bactérienne : proposition d'un bioréacteur de désulfuration, optimisation des deux opérations du procédé." Paris 6, 1986. http://www.theses.fr/1986PA066309.
Full textAbstract:
Conception et réalisation d'un pilote de laboratoire de 20 litres, garni de supports bactériens en PVC, en vue de produire du biogaz (avec des teneurs en H2s au moins égales à 0,3%. ) Emploi de déchets agro-alimentaires riches en sulfates. L'automatisation est réalisée en connectant des capteurs et des actionneurs à un ordinateur industriel. Le biogaz produit passe en continu et en ligne dans un bioréacteur de désulfuration en molécules organiques. Proposition d'un modèle intégrant les différentes vitesses de réaction. Perspectives de synthèse de molécules à haute valeur ajoutée.
38
Bokretsion, Rahel Girmai. "On-line process control in pharmaceutical industry." Diss., 2003. http://hdl.handle.net/2263/27182.
Full text
39
Der-Fang, Yang, and 楊德芳. "On-Line Process Quality Control Rules' Establishment and Practicability." Thesis, 1995. http://ndltd.ncl.edu.tw/handle/15536472909833206453.
Full textAbstract:
碩士東海大學
工業工程研究所
83
Shewhart(1931)提出品質管制圖及檢定規則,此後發展的管制圖大都依循 此種判斷方式,這種屬於離線方式的制程診斷主要是針對人工檢驗而設計 的,且目前所發展的品質管制專家系統亦是延用傳統的管制法則. 但由於 科技的進步與電腦技術的發展,檢驗自動化已成為趨勢, 而傳統管制法則 卻不能滿足此種檢驗的方式. 因此發展線上製程品質管制法則是必要的, 也就是希望藉由此種法則能夠立即偵測製程狀況,避免下一個不良品的產 生,以解決傳統管制圖反應延遲的問題.本研究主要是假設當製程穩定時, 觀察值相互獨立且服從常態分配, 利用統計技巧中母體比率的假設檢定來 發展線上製程品質管制法則,期望錯誤警報率0.005. 最後運用影像處理的 技術作實例分析. The quality control chart and inspection rules were proposed by Shewhart in 1931. The off-line process control diagnosis is mainly designed for artificial inspection, and development of expert system of quality control at present continuse following the tradition process control rules. For progress in science and computer technology makes the automated inspection be available. However, the tradition process control rules can't fit the style of the inspection. Therefore, it is necessary to develop the "On-Line process Quality Control Rules". That is, expecting to detect the condition of process immediately to avoid the next defective and solving the delay of reaction of the traditional process control charts. This research, which assumes when the process stable, and the data is independent and identically distributed (iid). By using the hypothesis testing of the multi-population proportion, develops the "On- Line Process Quality Control Rules" to expect the probability of false alarm to be 0.005. Finally, an actual example by using the technique of image process is made to reveal the practical way of those rules developed in this research.
40
GUO, JUN-MING, and 郭俊明. "On-line identification and PID control for a heating process." Thesis, 1986. http://ndltd.ncl.edu.tw/handle/22544600254615067128.
Full text
41
Wang, Chung-Cheng, and 王忠誠. "Research on Lithographic In-line Process Control from Scatterometry Measurements." Thesis, 2009. http://ndltd.ncl.edu.tw/handle/56353516546764590624.
Full textAbstract:
碩士國立交通大學
工學院碩士在職專班半導體材料與製程設備組
97
As lithography technology keep shrinking CD size, the accuracy of CD measurement and process control become more and more important. Improve within wafer CDU also can improve product yield. After photo-resist from DNQ/Nolvolak resist change to CAR (Chemical Amplified Resist), PEB(Post Exposure bake) become the key factor of litho CDU control. For nano-node lithography, traditional CD measurement methodology looks meet the resolution limit of semiconductor component. CD-SEM is from top to bottom measurement, can not provide detail profile information. Unless use cross section method, or can not show the depth data. But this way need to damage wafer also waste a lot time. As the semiconductor technology development, optical scatterometry will be next generation of CD measurement. Currently, the key point of PEB temperature is hot plate thermal controller. But no matter how accurate of hot plate temperature control is, seem still not meet the CDU requirement of nano-lithography. Someone tried to calibrate the PEB temperature by using development CD, in order to make the within wafer CDU better. Whatever, these methods are using SEM by CD measurement. Still need lot measurement point and time for data accuracy. As the optical CD technology improve, the features of accuracy、quickly and more profile information maybe provide the new direction of PEB temperature calibration by development CD. In this paper, the feasibility of improving the within wafer CDU will be discuss by using integrated scatterometry. The target of the works are optical CD match with SEM and PEB temperature calibration by integrated scatterometry for within wafer CDU best.
42
Kao, Chuang-Yih, and 高創義. "The Study of Exhause Dyeing Process by On-Line Control System." Thesis, 1996. http://ndltd.ncl.edu.tw/handle/99003966127392746258.
Full text
43
Chang, Jen-Ta, and 張仁達. "On-Line Process Control and Diagnostic System Using Neural Network Approach." Thesis, 2004. http://ndltd.ncl.edu.tw/handle/28724131158895491459.
Full textAbstract:
碩士國立雲林科技大學
工業工程與管理研究所碩士班
92
In this research, an on-line process control and diagnostic system with combining the concept of Statistical Process Control (SPC) and Engineering Process Control (EPC) is developed. A back-propagation neural network is applied to construct the control and diagnostic system for the multiple-input and multiple-output (MIMO) system. During the steady process output, the neural network controller with disturbed output prediction is used to adjust process parameters to reduce the output errors that come from the non-assignable disturbance. In addition, because of the capability of parameter regulation, this controller is also introduced to search the best process parameters and associated outputs. When the process variances are caused by the assignable symptoms and the system becomes unstable, first, the fuzzy-rule based system identifies the abnormal conditions from the control chart of SPC, and then the neural network diagnostic system finds the causes for the problems. Finally, a simulation study based on Chemical Mechanical Polishing (CMP) process is used to verify the proposed concept. The simulation results show that the proposed neural network control and diagnostic system can control the disturbed output and diagnose the problem effectively.
44
Hong-Ming, Tsai, and 蔡宏明. "Applications of an On-line Learning Neuro-Controller to Process Control." Thesis, 1998. http://ndltd.ncl.edu.tw/handle/40013948038619157456.
Full textAbstract:
碩士大葉大學
食品工程研究所
86
A back-propagation (BP) neural network is a multi-layer feed- forwardneural network. The BP is typical and most frequently used in currentlearning models. An on-line learning neural network controller based on aback-propagation neural network has been used in this simulation. Thiscontroller can breakthrough the bottleneck of a traditionalneuro-controller, which can learn on-line and take a control action at thesame time. Furthermore, the controller can merge the learning process with areal control, eliminating the need to training it before use. The weights ofa neural network are obtained from a random number generator. Thus, settingweights is not a problem. In this study, the on-line learning neural network controller has beenapplied to a CSTR system. The effect of learning rates on controllability ofan on-line learning neuro-controller has been studied in this simulation.The computer simulation of a CSTR system with two control loops has beencarried out. The interaction between the two loops is also discussed.Simulation results show that the neuro-controller with fixed learning ratescannot control the CSTR at satisfactory standers. Therefore, a linearexpansion method and Newton likely method are then used to obtain optimal,dynamic learning rates. Simulation results show that good controlperformance is obtained if an on-line learning neuro-controllers combinedwith optimal, dynamic learning rates. A neural network needs to be trained before use. Training a neuralnetwork takes a lot of time, and therefore limits the application of theneural network. If a neural network can take training and perform a controlaction at the same time, then the application of a neural network becomesmore convenient. Today''s personal computers have improved calculationability, which can be used to train neuro-controllers to predict a nearfuture process state. This ability is very important to a real-timesimulation. Actually, the neuro- controllers play two roles, includingfeed-froward (prediction) and feed-back (control).
45
Chen, Jin Zhang, and 陳金璋. "On-line tool failure detection for control of the milling process." Thesis, 1996. http://ndltd.ncl.edu.tw/handle/80465268854093035661.
Full text
46
Tsai, Po-Feng, and 蔡柏灃. "NLPCA and Quality Control of Autoclave Process batches on line operations." Thesis, 1998. http://ndltd.ncl.edu.tw/handle/82442055359539131549.
Full text
47
Chin, Po-Hsin, and 金柏馨. "The Off-line and On-line Control System for Automated Cementing Process of Shoe Manufacturing Industry." Thesis, 2015. http://ndltd.ncl.edu.tw/handle/37939953369011572115.
Full textAbstract:
碩士國立交通大學
工業工程與管理系所
103
The cementing process of shoe manufacturing industry mainly relies on labor. However, the labor cost has increased rapidly over the past few years. Therefore, shoe industry need to install automation system for cementing process to reduce the labor cost. Shoe industry usually evaluates the quality of cementing process through conducting tensile test on several locations of the sole and multiple readings are obtained from various locations for each sole. Therefore, when finding the optimal setting of parameter for the automation equipment, the mean and variation of tensile strength need to be considered simultaneously. Also, the target value of tensile strength in some part of the sole is higher than that of other parts. Furthermore, keeping the cost down is also important for adopting automation system. After finding the optimal parameter settings, the next stage is mass production. Another objective of this study is to construct an on-line monitoring system for the automatic cementing process. Because there are multiple readings per sole, variation of tensile strength may come from several sources. It is necessary to monitor the variation within one sole, variation among soles of the same batch and variation among different batches of soles, respectively. In other words, a three-stage controlling system used to effectively find the source of variations. Thus , one of the objective of this study is to find the optimal parameter setting for the automation system of cementing process using design of experiments(DOE) method. Finally, a real case provided by a shoe manufacturing company in Taiwan to demonstrate the effectiveness of the proposed method.
48
Wong, Wai-leung. "Incorporating variability into on-line feedback control for improved process stability and productivity." 1987. http://catalog.hathitrust.org/api/volumes/oclc/17543611.html.
Full textAbstract:
Thesis (M.S.)--University of Wisconsin--Madison, 1987.Typescript. eContent provider-neutral record in process. Description based on print version record. Includes bibliographical references (leaves 95-96).
49
GOTO, MASAFUMI. "ON-LINE ESTIMATION OF OXYGEN UPTAKE RATE FOR THE ACTIVATED SLUDGE PROCESS (CONTROL, SIMULATION, TRANSFER)." Thesis, 1985. http://hdl.handle.net/1911/15901.
Full textAbstract:
The purpose of this study was to develop a technique for on-line estimation of oxygen uptake rate (OUR) and to investigate the use of OUR in monitoring and control of the activated sludge process.
A mathematical expression for the relationship between the plant specific oxygen transfer coefficient (K) and air flow rate was necessary for the estimation of OUR. Three different least squares methods, a direct linear method, a discrete linear method, and a non-linear method, were investigated as estimation techniques for K.
Data from a full-scale, instrumented wastewater treatment plant were collected in order to evaluate the usefulness of each of the three methods. It was found that the direct linear method with data smoothing, although mathematically less rigorous than the other two methods, yielded a satisfactory estimate and was preferred over the non-linear method because of its computational simplicity. Difficulties were encountered with the discrete linear method because of disturbances in flow rate at the plant due to the on-off operation of several influent pumps.
A comparison between the measured and estimated OUR showed similar trends indicating that the estimated OUR could be used in control. The use of OUR for control purposes was investigated by means of computer simulations using an existing mathematical model modified to simulate conditions at the full-scale plant. Control strategies, using OUR estimates, for air flow rate and step feed manipulation were developed and demonstrated to be effective.
Because OUR is a better indicator of sludge activity than dissolved oxygen, the estimated OUR may also be used to monitor process conditions and obtain an estimate of the active fraction in the sludge. The latter may enable improved control strategies for sludge age to be developed.
50
Ho, Chien-Da, and 何建鐽. "Development of an On-Line Monitoring Control System for Abnormal Machining and a Machining Process Improvement System." Thesis, 2011. http://ndltd.ncl.edu.tw/handle/um3sfb.
Full textAbstract:
碩士中原大學
機械工程研究所
99
By using signal processing and diagnosis technology and ICT technology, the main purpose of this study is to develop an on-line intelligent monitoring and control system of abnormal machining and a machining process improvement system for CNC machine tools. The system can be used to enhance the machining accuracy, improve yield rate of production, and optimize the machining parameters. The system is designed to monitor/control three type of abnormal machining phenomenon: chatter, built-up-edge, and cutter breakage. The system composes two units: (1) on-site monitoring unit; (2) process optimization unit. To increase diagnosis efficiency and reliability, the on-line monitoring system uses signal variation characteristics analysis and Fast Fourier Transfer instead of complex single transfer methods for signal analysis. Besides, Fanuc Open CNC Api Spec is used to develop a program which can on-line extract the instant NC codes and machine state parameters from the CNC controller as reference to avoid fail diagnosis. The system can provide early warning for and control for abnormal machining to improve machining accuracy and protect the machine and cutter. The experimental results have shown that the chatter can be diagnosed within 1 second and suppressed within 3 seconds, and the built-up-edge and cutter breakage can be diagnosed within 3 seconds and controlled within 5 seconds. In addition, though analyzing the monitoring data and computation with process optimizing algorithm the process optimization unit can recommend the machining parameters for better machining efficiency or better machining accuracy.