Дисертації з теми "Multirole"

Aujourd'hui, les opportunités pour les étudiants de s'immerger dans des situations professionnelles concrètes restent limitées, se résumant souvent à des projets de groupe ou des stages. Cette thèse se situe dans le domaine des Environnements Informatique pour l'Apprentissage Humain (EIAH) et explore la conception de jeux sérieux pour l'apprentissage, ou learning games, dont l'usage comme outil pédagogique peut permettre de combler cette lacune. Nous nous concentrons sur le développement de jeux de rôles (Role Learning Games - RLG), destinés à simuler des environnements professionnels multi-acteurs. Contrairement aux jeux sérieux traditionnels orientés vers le jeu à un seul joueur, les RLG visent à reproduire la complexité des interactions professionnelles en encourageant la coopération entre différents rôles, chacun ayant des objectifs et des activités distincts. Notre problématique est donc : comment modéliser des scénarios avec interactions entre joueurs pour des learning games asymétriques ? L'état de l'art révèle une carence dans les méthodes et outils existants pour la création de tels jeux. Face à ce constat, nous proposons un modèle pour la conception de RLG asymétriques, permettant des interactions entre les joueurs. Suite à cela, nous avons réifié ce modèle au travers de deux outils, l'un tangible sous forme d'un puzzle, l'autre numérique et disponible en ligne. Enfin, les résultats obtenus au cours de des expérimentations sont présentés, démontrant l'intérêt du modèle et de ces outils quant à l'accompagnement de l'enseignant durant la phase de scénarisation. La conclusion souligne l'apport de cette recherche, tout en ouvrant la voie à des perspectives à venir.Notre domaine de recherche est celui des EIAH (Environnements informatiques pour l'apprentissage humain) avec une orientation informatique. Nous nous plaçons dans le paradigme épistémologique du constructivisme pragmatique. Cette recherche a été guidée par le Design Based Research ou DBR, et ont suivi la méthode THEDRE : Traceable Human Experiment Design Research de Nadine Mandran. Pour répondre à notre problématique, nous avons mené plusieurs expérimentations qualitatives. Les utilisateurs finaux impliqués dans les expérimentations étaient en très grande majorité des enseignants (de collège, lycée ou école d'ingénieurs) et des ingénieurs pédagogiques. Suite aux expérimentations, nous avons mené des focus groups ou entretiens individuels. Les échanges enregistrés lors des sessions ont fait l'objet d'une analyse thématique. L'ensemble des 11 expérimentations conduites avec 54 participants, principalement des enseignants et ingénieurs pédagogiques, a permis de développer un modèle innovant pour la création de scénarios de RLG. Ce modèle favorise la mise en œuvre de la coopération entre les participants du futur RLG, chacun jouant des rôles distincts avec des objectifs pédagogiques spécifiques. Pour évaluer notre modèle, nous avons créé deux outils complémentaires : un kit tangible (RLG Kit) et une application numérique (RLG Maker). Grâce à ces outils, deux RLG ont été joués en classe avec une centaine d'élèvesau total. En outre, un système de vérification des scénarios (RLG Checker) a été développé pour assister les concepteurs dans la scénarisation, en s'assurant que les rôles peuvent coopérer efficacement. En résumé, cette thèse contribue à l'avancement des EIAH en proposant un modèle deconception de scénarios pour les Role Learning Games, complété par des outils de scénarisation polyvalents et un système de vérification. Les premières applications de ces outils ont prouvé leur efficacité dans la création de jeux éducatifs, recevant un accueil positif des enseignants. Il est actuellement possible d'étendre leur utilisation au-delà des contextes pédagogiques et nous travaillons sur une intégration des scénarios sous Unity pour générer des jeux en 3D grâce au RLG Maker et un modèle réalisé pour ce moteur de jeux
Today, opportunities for students to immerse themselves in concrete professional situations remain limited, often confined to group projects or internships. This thesis is situated in the field of Technology Enhanced Learning (TEL) and explores the design of serious games for learning, known as learning games, whose use as educational tools can help fill this gap. We focus on the development of Role Learning Games (RLG), intended to simulate multi-actor professional environments. Unlike traditional serious games oriented toward single-player gameplay, RLGs aim to replicate the complexity of professional interactions by encouraging cooperation between different roles, each with distinct objectives and activities. Our research problem thus deals with modeling scenarios with player interaction for asymmetrical learning games. The state of the art reveals a lack of existing methods and tools for creating such games. In response, we propose a model for designing asymmetric RLGs allowing interactions between players. Subsequently, we reified this model through two tools, one tangible in the form of a puzzle, the other digital and available online. Finally, the results obtained from the experiments are presented, demonstrating the interest of the model and these tools in supporting the teacher during the scenario development phase. The conclusion highlights the contribution of this research, while opening the way for future perspectives. Our research domain is TEL (Technology Enhanced Learning). We operate within the epistemological paradigm of pragmatic constructivism. This research was guided by Design Based Research (DBR) and we followed the THEDRE method: Traceable Human Experiment Design Research by Nadine Mandran. To address our research question, we conducted several qualitative experiments. The end-users involved in the experiments were predominantly teachers (from middle schools, high schools, or engineering schools) and instructional designers. Following the experiments, we conducted focus groups or individual interviews. The recorded exchanges during the sessions were subjected to thematic analysis.A total of 11 experiments conducted with 54 participants, primarily teachers and instructional designers, allowed the development of an innovative model for creating RLG scenarios. This model promotes the implementation of cooperation between the participants of the future RLG, each playing distinct roles with specific educational objectives. To evaluate our model, we created two complementary tools: a tangible kit (RLG Kit) and a digital application (RLG Maker). With these tools, two RLGs were played in class with around a hundred students in total. Additionally, a scenario verification system (RLG Checker) was developed to assist designers in scenario development, ensuring that roles can cooperate effectively.In summary, this thesis advances TEL by proposing a scenario design model for Role Learning Games, complemented by versatile scenario design tools and a verification system. The first applications of these tools have proven their effectiveness in creating educational games, receiving positive feedback from teachers. It is currently possible to extend their use beyond educational contexts, and we are working on integrating scenarios into a Unity template to generate 3D games using the RLG Maker

Thesis (Degree of Electrical Engineer) Naval Postgraduate School, Dec. 2000.
Thesis advisor, Charles W. Therrien. "December 2000." Includes bibliographical references (p. 91-92). Also available in print.

A rapidly growing interest in third generation Personal Communication Networks has underlined the importance of wireless multimedia systems that can support voice, video, images, files, or any combinations thereof. In order to ensure satisfactory quality of service (QoS) for individual multimedia traffic, a new analysis with a new user model based on circuit-switched direct-sequence Code Division Multiple Access (DS-CDMA) system is presented. The new user model is introduced to allow users to transmit data at multiple bit rates and to switch to other bit rates at any time. To facilitate performance analysis, each traffic type with rate s is assumed a probability ps for user k. To ensure satisfactory QoS, a new power control scheme is further proposed for a multimedia circuit-switched DS-CDMA system. Specifically, a new closed-form power control function is introduced to ensure quality of service for each traffic type and to achieve a better overall throughput at the same time. Central to the new closed-form power control function is a parameter called the traffic exponent. By introducing this parameter, the difficulty in obtaining an optimal closed-form power control function is reduced which simplifies the information feedback process from the base station to the mobile stations.

Approved for public release; distribution in unlimited.
This thesis addresses the problem of estimating a random process from two observed signals sampled at different rates. The case where the low-rate observation has a higher signal-to- noise ratio than the high-rate observation is addressed. Both adaptive and non-adaptive filtering techniques are explored. For the non-adaptive case, a multirate version of the Wiener-Hopf optimal filter is used for estimation. Three forms of the filter are described. It is shown that using both observations with this filter achieves a lower mean-squared error than using either sequence alone. Furthermore, the amount of training data to solve for the filter weights is comparable to that needed when using either sequence alone. For the adaptive case, a multirate version of the LMS adaptive algorithm is developed. Both narrowband and broadband interference are removed using the algorithm in an adaptive noise cancellation scheme. The ability to remove interference at the high rate using observations taken at the low rate without the high-rate observations is demonstrated.

Thesis (M.S. in Electrical Engineering)--Naval Postgraduate School, September 2003.
Thesis advisor(s): Charles W. Therrien, Roberto Cristi. Includes bibliographical references (p. 45). Also available online.

Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1991.
Title as it appears in the Feb. 1991 M.I.T. Graduate List: Robust control for asynchronous multitrate sampled systems.
Includes bibliographical references (leaves 91-94).
by Joël Rosenberg.
M.S.

Doutoramento em Engenharia Electrotécnica
Esta tese insere-se na área da simulação de circuitos de RF e microondas, e visa o estudo de ferramentas computacionais inovadoras que consigam simular, de forma eficiente, circuitos não lineares e muito heterogéneos, contendo uma estrutura combinada de blocos analógicos de RF e de banda base e blocos digitais, a operar em múltiplas escalas de tempo. Os métodos numéricos propostos nesta tese baseiam-se em estratégias multi-dimensionais, as quais usam múltiplas variáveis temporais definidas em domínios de tempo deformados e não deformados, para lidar, de forma eficaz, com as disparidades existentes entre as diversas escalas de tempo. De modo a poder tirar proveito dos diferentes ritmos de evolução temporal existentes entre correntes e tensões com variação muito rápida (variáveis de estado activas) e correntes e tensões com variação lenta (variáveis de estado latentes), são utilizadas algumas técnicas numéricas avançadas para operar dentro dos espaços multi-dimensionais, como, por exemplo, os algoritmos multi-ritmo de Runge-Kutta, ou o método das linhas. São também apresentadas algumas estratégias de partição dos circuitos, as quais permitem dividir um circuito em sub-circuitos de uma forma completamente automática, em função dos ritmos de evolução das suas variáveis de estado. Para problemas acentuadamente não lineares, são propostos vários métodos inovadores de simulação a operar estritamente no domínio do tempo. Para problemas com não linearidades moderadas é proposto um novo método híbrido frequência-tempo, baseado numa combinação entre a integração passo a passo unidimensional e o método seguidor de envolvente com balanço harmónico. O desempenho dos métodos é testado na simulação de alguns exemplos ilustrativos, com resultados bastante promissores. Uma análise comparativa entre os métodos agora propostos e os métodos actualmente existentes para simulação RF, revela ganhos consideráveis em termos de rapidez de computação.
This thesis belongs to the field of RF and microwave circuit simulation, and is intended to discuss some innovative computer-aided design tools especially conceived for the efficient numerical simulation of highly heterogeneous nonlinear wireless communication circuits, combining RF and baseband analog and digital circuitry, operating in multiple time scales. The numerical methods proposed in this thesis are based on multivariate strategies, which use multiple time variables defined in warped and unwarped time domains, for efficiently dealing with the time-scale disparities. In order to benefit from the different rates of variation of slowly varying (latent) and fast-varying (active) currents and voltages (circuits’ state variables), several advanced numerical techniques, such as modern multirate Runge-Kutta algorithms, or the mathematical method of lines, are proposed to operate within the multivariate frameworks. Diverse partitioning strategies are also introduced, which allow the simulator to automatically split the circuits into sub-circuits according to the different time rates of change of their state variables. Novel purely time-domain techniques are addressed for the numerical simulation of circuits presenting strong nonlinearities, while a mixed frequency-time engine, based on a combination of univariate time-step integration with multitime envelope transient harmonic balance, is discussed for circuits operating under moderately nonlinear regimes. Tests performed in illustrative circuit examples with the newly proposed methods revealed very promising results. Indeed, compared to previously available RF tools, significant gains in simulation speed are reported.

The company Spotify enables hardware companies to produce sound systems that are capable of playing music via Spotify’s music service. The sound systems are then sold to consumers. In order to guarantee a certain level of quality, the products need to be certified, mainly in regards to performance. This is ideally performed automatically, with the aid of some computer software. In this project, a tool was designed and developed to measure the performance of sound systems, especially those consisting of several loudspeakers, connected via a local area network. The measurements are taken by ordering the sound system to play test tones. The test tones are then recorded, identified and timed by a computer program, by applying a cross-correlation technique, which is otherwise used to perform echo-cancellation among other things. The tool provides improvements and extends the functionality of an existing tool used currently by Spotify.
Företaget Spotify låter företag inom hårdvarubranschen tillverka ljudsystem, som kan spela musik via Spotifys musiktjänst. Ljudsystemen säljs sedan till konsumenter. För att kunna garantera att produkterna möter en viss kvalitétsnivå behöver man certifiera produkterna, främst med avseende på deras prestanda. Detta görs lämpligast automatiskt, med ett datorverktyg. I det här projektet designades och utvecklades ett verktyg för att automatiskt mäta prestandan hos ljudsystem, speciellt sådana som består av flera högtalare, sammanlänkade via ett lokalt datornätverk. Mätningen går till så att ljudsystemet beordras att spela upp testtoner, som sedan spelas in, identifieras och tidmäts av ett datorprogram, med hjälp av krosskorrelationsteknik, som annars används i till exempel ekokancellering. Verktyget erbjuder förbättringar och kompletterar funktionaliteten hos ett befintligt verktyg som för närvarande används på Spotify.

This report describes the design and implementation of a fixed audio equalizer based on a scheme where parts of the signal spectrum are downsampled and treated differently for the purpose of reducing the computational complexity and memory requirements. The primary focus has been on finding a way of taking an equalizer based on a simple minimum-phase FIR filter and transform it to the new type of equalizer. To achieve this, a number of undesireable effects such as aliasing distortion and upsampling imaging had to be considered and dealt with. In order to achieve a good amplitude response of the system, optimization procedures were used. As part of the thesis, a cost-effective implementation of the filter has been made for an FPGA, in order to verify that the scheme is indeed usable for equalizing an audio signal.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Civil and Environmental Engineering, 2004.
Includes bibliographical references (p. 219-230) and index.
Wavelets and filter banks have revolutionized signal processing with their ability to process data at multiple temporal and spatial resolutions. Fundamentally, continuous-time wavelets are governed by discrete-time filter banks with properties such as perfect reconstruction, linear phase and regularity. In this thesis, we study multi-channel filter bank factorization and parameterization strategies, which facilitate designs with specified properties that are enforced by the actual factorization structure. For M-channel filter banks (M =/> 2), we develop a complete factorization, M-channel lifting factorization, using simple ladder-like structures as predictions between channels to provide robust and efficient implementation; perfect reconstruction is structurally enforced, even under finite precision arithmetic and quantization of lifting coefficients. With lifting, optimal low-complexity integer wavelet transforms can thus be designed using a simple and fast algorithm that incorporates prescribed limits on hardware operations for power-constrained environments. As filter bank regularity is important for a variety of reasons, an aspect of particular interest is the structural imposition of regularity onto factorizations based on the dyadic form uvt. We derive the corresponding structural conditions for regularity, for which M-channel lifting factorization provides an essential parameterization. As a result, we are able to design filter banks that are exactly regular and amenable to fast implementations with perfect reconstruction, regardless of the choice of free parameters and possible finite precision effects. Further constraining u = v ensures regular orthogonal filter banks,
(cont.) whereas a special dyadic form is developed that guarantees linear phase. We achieve superior coding gains within 0.1% of the optimum, and benchmarks conducted on image compression applications show clear improvements in perceptual and objective performance. We also consider the problem of completing an M-channel filter bank, given only its scaling filter. M-channel lifting factorization can efficiently complete such biorthogonal filter banks. On the other hand, an improved scheme for completing paraunitary filter banks is made possible by a novel order-one factorization which allows greater design flexibility, resulting in improved frequency selectivity and energy compaction over existing state of the art methods. In a dual setting, the technique can be applied to transmultiplexer design to achieve higher-rate data transmissions.
by Ying-Jui Chen.
Ph.D.

Reactive transport modeling is a methodological tool to study the coupled physical, chemical and biological processes in Earth system. It is complex not only because of the nature of the equations, but also because of the effects of the porous medium heterogeneity on reactive transport. This thesis aims to deepen the understanding of reactive transport processes in order to explain the biochemical degradation process in porous media, with special emphasis on the role of biofilm and its growth. First, we propose a general and efficient numerical solution of reactive transport in multicontinuum media using Multirate Mass Transfer (MRMT) approach. To overcome the non-linearity of the problem, induced by non-linear kinetics, we use the Newton-Raphson method to get the global solution. We solve the system of equations in block form, which allow us to reduce the unknowns to those of mobile zones and to, thus improving efficiency. The solution is validated by comparison with analytical solution for linear kinetics. The code is developed in Object-oriented way, which enables the code reusability and data polymorphism. Second, we investigate the conditions for chemical localization (i.e., the occurrence of reactions that would not be possible in single continuum media). To this end, we write the multicontinuum transport equations in dimensionless form to find that reactive transport in multicontinuum media is governed by three characteristic times: the distribution of residence times in immobile zones, and the characteristic reaction and transport times. To study the interplay between these three characteristic times, we simulate three chemical systems: conservative, single reaction and sequential reaction. Results demonstrate that reactions driven by species that result from previous reactions will localize in immobile zones whose residence time is comparable to reaction times. Furthermore, immobile zones with residence times much smaller than those for transport can be lumped together (assuming that very fast reactions are assumed in equilibrium), which greatly reduces computations. Third, we perform simulations of reactive transport incorporating biochemical reactions that not only oxidize organic carbon, but also produce biomass, thus causing biofilm growth. Biofilm growth is known to cause clogging (i.e., reduction of permeability), which has concentrated most research on the topic. But it also causes a significant change in the pore space geometry and connectivity, which leads to not only an overall increase in mean residence time in immobile regions, but also on its distribution. As discussed above, this is critical to (bio)chemical localization, especially considering that microbial mediated reactions tend to concentrate in biofilms. We propose a model for the evolution of residence time distribution in immobile zones in response to biofilm growth. We test this model by comparison with laboratory experiments extracted from the literature, where tracer tests have been performed at various stages of growth. Results show that the dynamic MRMT model is capable of reproducing the salient features of these experiments.
El modelado de transporte reactivo es una herramienta metodológica para estudiar los procesos físicos, químicos y biológicos acoplados en el sistema terrestre. Es complejo no solo por la naturaleza de las ecuaciones, sino también por los efectos de la heterogeneidad del medio poroso sobre el transporte reactivo. Esta tesis tiene como objetivo profundizar en el conocimiento de los procesos de transporte reactivo para explicar el proceso de degradación bioquímica en medios porosos, con especial énfasis en el papel del biofilm y su crecimiento. En primer lugar, proponemos una solución numérica general y eficiente de transporte reactivo en medios multicontinuum utilizando el enfoque de Transferencia de Masa Multivelocidad (MRMT). Para superar la no linealidad del problema, inducida por una cinética no lineal, utilizamos el método de Newton-Raphson para obtener la solución global. Resolvemos el sistema de ecuaciones en forma de bloque, lo que nos permite reducir las incógnitas a las de zonas móviles y a, mejorando así la eficiencia. La solución se valida por comparación con la solución analítica para cinética lineal. El código se desarrolla de forma orientada a objetos, lo que permite la reutilización del código y el polimorfismo de los datos. En segundo lugar, investigamos las condiciones para la localización química (es decir, la ocurrencia de reacciones que no serían posibles en un medio continuo único). Con este fin, escribimos las ecuaciones de transporte multicontinuum en forma adimensional para encontrar que el transporte reactivo en medios multicontinuum se rige por tres tiempos característicos: la distribución de los tiempos de residencia en zonas inmóviles, y los tiempos de reacción y transporte característicos. Para estudiar la interacción entre estos tres tiempos característicos, simulamos tres sistemas químicos: conservador, reacción única y reacción secuencial. Los resultados demuestran que las reacciones impulsadas por especies que resultan de reacciones previas se localizarán en zonas inmóviles cuyo tiempo de residencia es comparable a los tiempos de reacción. Además, las zonas inmóviles con tiempos de residencia mucho más pequeños que los de transporte pueden agruparse (asumiendo que se asumen reacciones muy rápidas en equilibrio), lo que reduce en gran medida los cálculos. En tercer lugar, realizamos simulaciones de transporte reactivo incorporando reacciones bioquímicas que no solo oxidan el carbono orgánico, sino que también producen biomasa, lo que provoca el crecimiento de biopelículas. Se sabe que el crecimiento de biopelículas causa obstrucciones (es decir, reducción de la permeabilidad), lo que ha concentrado la mayor parte de la investigación sobre el tema. Pero también provoca un cambio significativo en la geometría del espacio poroso y la conectividad, lo que conduce no solo a un aumento general del tiempo medio de residencia en las regiones inmóviles, sino también a su distribución. Como se discutió anteriormente, esto es crítico para la localización (bio) química, especialmente considerando que las reacciones mediadas por microbios tienden a concentrarse en biopelículas. Proponemos un modelo para la evolución de la distribución del tiempo de residencia en zonas inmóviles en respuesta al crecimiento de biopelículas. Probamos este modelo comparándolo con experimentos de laboratorio extraídos de la literatura, donde se han realizado pruebas de trazadores en varias etapas de crecimiento. Los resultados muestran que el modelo MRMT dinámico es capaz de reproducir las características más destacadas de estos experimentos.
La modelització del transport reactiu és una eina metodològica per estudiar els processos físics, químics i biològics acoblats en el sistema terrestre. És complex no només per la naturalesa de les equacions, sinó també pels efectes de l’heterogeneïtat mitjana porosa sobre el transport reactiu. Aquesta tesi té com a objectiu aprofundir en la comprensió dels processos de transport reactiu per explicar el procés de degradació bioquímica en medis porosos, amb especial èmfasi en el paper del biofilm i el seu creixement. En primer lloc, proposem una solució numèrica general i eficient de transport reactiu en medis multicontinuos mitjançant l'enfocament de transferència de massa multirata (MRMT). Per superar la no linealitat del problema, induïda per cinètiques no lineals, fem servir el mètode de Newton-Raphson per obtenir la solució global. Resolem el sistema d’equacions en forma de blocs, que ens permeten reduir les incògnites a les de les zones mòbils i a, millorant així l’eficiència. La solució es valida comparant-la amb una solució analítica de cinètica lineal. El codi es desenvolupa de manera orientada a objectes, que permet la reutilització del codi i el polimorfisme de dades. En segon lloc, investigem les condicions per a la localització química (és a dir, l’aparició de reaccions que no serien possibles en un mitjà continu). Amb aquest objectiu, escrivim les equacions de transport multicontinu en forma adimensional per trobar que el transport reactiu en mitjans multicontinuos es regeix per tres temps característics: la distribució dels temps de residència en zones immòbils i els temps de reacció i transport característics. Per estudiar la interacció entre aquests tres temps característics, simulem tres sistemes químics: conservador, de reacció única i reacció seqüencial. Els resultats demostren que les reaccions impulsades per espècies que resulten de reaccions anteriors es localitzaran en zones immòbils el temps de residència és comparable als temps de reacció. A més, les zones immòbils amb temps de residència molt menors que les del transport es poden agrupar (suposant que s’assumeixen reaccions molt ràpides en equilibri), cosa que redueix considerablement els càlculs. En tercer lloc, realitzem simulacions de transport reactiu que incorporen reaccions bioquímiques que no només oxiden el carboni orgànic, sinó que també produeixen biomassa, provocant així el creixement del biofilm. Se sap que el creixement del biofilm causa l’obstrucció (és a dir, la reducció de la permeabilitat), que ha concentrat la majoria de les investigacions sobre el tema. Però també provoca un canvi significatiu en la geometria i la connectivitat de l’espai dels porus, que condueix no només a un augment global del temps mitjà de residència en regions immòbils, sinó també a la seva distribució. Com s’ha comentat anteriorment, això és fonamental per a la localització (bio) química, sobretot tenint en compte que les reaccions mediàtiques microbianes tendeixen a concentrar-se en biofilms. Proposem un model per a l’evolució de la distribució del temps de residència en zones immòbils en resposta al creixement del biofilm. Provem aquest model en comparació amb experiments de laboratori extrets de la literatura, on s'han realitzat proves de traçador en diverses etapes de creixement. Els resultats mostren que el model dinàmic MRMT és capaç de reproduir les característiques més destacades d’aquests experiments
(Xinès) 反应运移模型是地球系统中研究物理、化学和生物耦合过程的一种方法工具。反应运移是复杂的，不仅在于其控制方程本身的复杂性，而且受多孔介质非均质性的影响。本论文旨在加深对反应运移过程的理解，以解释多孔介质中的生物化学降解过程，尤其是生物膜的生长及其作用。 首先，我们利用多速率传质(MRMT)方法提出了多连续介质中反应运移的一种通用且高效的数值解。为克服非线性动力学反应导致问题的非线性，我们采用牛顿-拉夫森方法迭代求得全局解。利用分块的形式求解方程组，将未知量的个数减少为可移动区域的个数，从而提高计算效率。通过与线性动力学解析解的比较，验证了该方法的正确性。采用面向对象的方式开发程序，实现了代码的可重用性和数据多态性。 其次，我们研究了局部化学反应的条件，即在单一连续介质中不可能发生的反应的发生。为此，我们推导了多连续介质中的反应运移方程的无量纲形式，发现多连续介质中的反应运移受控于三个特征时间:不可移动区停留时间的分布，反应时间和运移时间。为研究这三种特征时间之间的相互作用，我们模拟了三种化学体系:保守反应、单反应和顺序反应。结果表明，由前一的反应导致的物种驱动的反应将定位于停留时间与反应时间相当的不可移动区域。此外，可将停留时间比运移时间小的多的不可移动区域视作可移动区域（假设非常快的反应处于平衡状态），从而很大的减少了计算。 最后，我们模拟了有微生物参与的反应运移，涉及的生物化学反应不仅能够氧化有机碳，而且能够为微生物的生长提供碳源，从而导致生物膜的生长。众所周知，生物膜的生长会导致生物堵塞（即渗透性的降的），大多研究致力于该方向。但生物膜的生长也会导致孔隙几何形状和连通性发生显著变化，不仅导致不可移动区域平均停留时间的增加，而且改变其分布。如上所述，这对于研究局部（生物）化学过程至关重要，特别是在考虑微生物介导的反应时，其往往集中在生物膜中。我们提出了一个停留时间在不可移动区域中随生物膜生长的演化模型。通过与文献中提取的实验数据进行对比，在生长的不同阶段进行示踪实验，来测试该模型。结果表明，该模型能够较好地再现实验的显著性特征。
Enginyeria del terreny

Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1993.
Includes bibliographical references (leaves 101-104).
by Jen Mei Chen.
Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1993.

Thesis (M. Phil.)--Hong Kong University of Science and Technology, 2002.
Includes bibliographical references (leaves 55-58). Also available in electronic version. Access restricted to campus users.

Le crescenti esigenze di capacità di connessione per le aree rurali, remote e anche urbane probabilmente aumenteranno il costo della pura copertura terrestre. In questo scenario le comunicazioni satellitari sono destinate a svolgere un ruolo significativo negli eco-sistemi 5/6G per fornire una copertura onnipresente, broadcast/multicast e l'impiego in situazioni di emergenza e disastri. A questo scopo, diverse costellazioni sono state proposte o sono già in fase di sviluppo sia per l'orbita geo-stazionaria (GEO) (ad esempio ViaSat-2, la rete Global Xpress (GX) di Inmarsat, la piattaforma EpicNG di Intelsat) e non-GEO (per esempio LeoSat, il sistema OneWeb, il sistema satellitare SpaceX Starlink). I transponder satellitari semi-trasparenti che si basano su processori digitali trasparenti (DTP) sono considerati come soluzioni interessanti per fornire un'ade-guata flessibilità rispetto agli standard in evoluzione e l'adattabilità rispetto ai modelli di traffico variabili nel tempo. Tuttavia, la loro progettazione hardware, presenta diverse criticità, specialmente quando si considerano grandi costellazioni di piccoli satelliti (es. LEO). In questo scenario è importante ricorrere all'uso di una procedura di progettazione a supporto, che muovendo dai requisiti di link-budget permette di effettuare un'accurata progettazione hardware basata su un trade-off tra prestazioni e complessità. Il presente lavoro propone un nuovo approccio di modellazione al fine di incorporare il consumo energetico delle architetture HW esplicitandolo nella fase concettuale della progettazione. Questa fase in ambito DSP spesso è legata alla forma aritmetica dell'algoritmo e come base vengono utilizzati i Signal Flow Graph (SFG). Il modello è stato utilizzato, in prima istanza, per studiare dal punto di vista energetico l'implementazione dell'intera catena del DTP su FPGA confrontando le varie soluzioni e individuando strategie più efficaci. In seconda istanza il metodo è stato poi utilizzato nel campo delle comunicazioni ottiche ed in particolare come base di progettazione per il modulatore e il demodulatore di un sistema di biotelemetria ottica basato su codifica multilivello ad impulsi corti. In questo caso l'analisi ha permesso di confrontare varie soluzioni per poi scegliere quella più adeguata dal punto di vista energetico durante la fase di progettazione concettuale di un caso reale. I risultati ottenuti per diverse configurazioni, requisiti e obiettivi di progettazione in entrambe i casi forniscono spunti promettenti nella prospettiva di formulare problemi di ottimizzazione. Infine viene definito un modello per la stima anticipata della potenza dinamica dei DTP, e non solo, implementati in FPGA. Il quadro teorico per i DTP è ulteriormente esteso per incorporare elementi legati alla tecnologia nel processo di analisi e progettazione. La complessità, che in lavori precedenti è stata espressa con blocchi elementari, è ulteriormente dettagliata attraverso l'espressione in termini di primitive hardware. Un particolare nuovo contributo è quello di stimare il consumo di energia e includerlo sia nella valutazione delle prestazioni che nell'approccio progettuale. Il modello anticipato di consumo energetico consente di ottenere espressioni per la complessità hardware e la stima della potenza che siano valide per gli FPGA attualmente disponibili in commercio, e in realizzazioni VLSI full-custom.

This thesis is concerned with the issues of design and optimisation of digital filters and multirate filter banks. The main focus and contribution of this thesis is to apply the genetic algorithm (GA) technique and to draw some comparison with the standard gradient and non-gradient based optimisation methods. The finite word length (FWL) constraint affects the accuracy of a real-time digital filter requency response. For the case of digital filters, this study is concerned with the optimisation of FWL coefficients using genetic algorithms. Some comparative study with the simple hill climber algorithms is also included. The outcome of this part of the study demonstrates a substantial improvement of the new results when compared with the simply rounded FWL coefficient frequency response. The FWL coefficient optimisation process developed in the earlier Chapters is extended to the field of multirate filter banks. All multirate filter banks suffer from the problems of amplitude, phase and aliasing errors and, therefore, constraints for perfect reconstruction (PR) of the input signal can be extensive. The problem, in general, is reduced to relaxing constraints at the expense of errors and finding methods for minimising the errors. Optimisation techniques are thus commonly used for the design and implementation of multirate filter banks. In this part of the study, GAs have been used in two distinct stages. Firstly, for the design optimisation so that the overall errors are minimised and secondly for FWL coefficient optimisation of digital filters that form the sub-band filters of the filter bank. This process leads to an optimal realisation of the filter bank that can be applied to specific applications such as telephony speech signal coding and compression. One example of the optimised QMF bank was tested on a real-time DSP target system and the results are reported. The multiple M-channel uniform and non-uniform filter banks have also been considered in this study for design optimisation. For a comparative study of the GA optimised results of the design stage of the filter bank, other standard methods such as the gradient based quasi-Newton and the non-gradient based downhill Simplex methods were also used. In general, the outcome of this part of study demonstrates that a hybrid approach of GA and standard method was the most efficient and effective process in generating the best results.

Programs are more distributed and concurrent today than ever before, and structural communications are at the core. Constructing and debugging such programs are hard due to the lack of formal specifications and verifications of concurrency. Recent advances in type systems allow us to specify the structures of communications as session types, thus enabling static type checking of the usages of communication channels against protocols. The soundness of session type systems implies communication fidelity and absence of deadlock. This work proposes to formalize multiparty dependent session types as an expressive and practical type discipline for enforcing communication protocols. The type system is formulated in the setting of multi-threaded λ-calculus with inspirations from multirole logic. It is sound, and it provides linearity and coherence guarantees entirely statically. The type system supports recursion and polymorphism. The formulation is particularly suitable for practical implementation, and this work provides such a runtime implementation.

碩士
國立臺灣大學
電信工程學研究所
103
With the increasing number of wireless communication devices, bandwidth becomes the precious resource. How to improve the bandwidth ef ciency is the important issue. Partial response (PR) signaling is bandlimited signaling. Because PR signaling can achieve maximum bandwidth ef ciency, PR signaling is widely applied to the wireless communication. However, because the impulse response of the bandlimited signal is in nite, it is not suitable for the burst transmission and low latency scenarios. In this thesis, we propose the multicarrier orthogonal multirate (MC-OMR) modulation which combines multicarrier modulation and OMR modulation which has been proposed. The MC-OMR waveform can have high energy compactness by designing the digital lter. By doing so, in practice, if the transmitted waveform is the truncation version of the MC-OMR waveform, the impact of the truncation on the data loss will be less signi cant. Therefore, the MC-OMR is suitable for the burst transmission. Moreover, according to the energy distribution property of MC-OMR signal, we propose the two truncation methods. In order to decrease the implementation complexity, we propose the ef cient structure of MC-OMR system.

碩士
明志科技大學
工程技術研究所
95
Due to the advance on digital computer and microprocessors , a great deal of efforts has been put into research in digital systems . In order to efficiently utilizes the computational capacity of the onboard computer and accommodate various sensors information available at multiple rates, multirate sampling techniques are applied to the control the systems in aerospace, automobile, hard disk driver, and DSP-based motor control fields. A systematic approach to the analysis of multirate systems is adopted first. The fuzzy logic controller design is then used to control multirate plants so that a better control effect can be obtained. The control rules and membership functions of fuzzy logic controllers are also optimized. Especially, the membership functions are optimized by genetic algorithm. The proposed intelligent control approach will be applied to disk drive servo control and automatic guided vehicle tracking control. Application results demonstrate the improved performance of the proposed approach for multirate control systems. Keywords：multirate systems, fuzzy logic controller, genetic algorithm, disk drive servo control, automatic guided vehicle.

Mui Sze Wai.
Thesis (M.Phil.)--Chinese University of Hong Kong, 2003.
Includes bibliographical references (leaves 62-65).
Abstracts in English and Chinese.
Chapter 1. --- Introduction --- p.1
Chapter 1.1 --- Evolution of Multirate Networks --- p.2
Chapter 1.2 --- Some Results from Previous Work --- p.2
Chapter 1.3 --- Multirate Traffic on Benes Network --- p.5
Chapter 1.4 --- Organization --- p.7
Chapter 2. --- Background Knowledge on Benes Network and Path Switching --- p.8
Chapter 2.1 --- Benes Network --- p.9
Chapter 2.1.1 --- Construction of Large Switching Fabrics --- p.9
Chapter 2.1.2 --- Routing in Benes Network --- p.11
Chapter 2.1.3 --- Performance when Operated as a Large Switch Fabric --- p.13
Chapter 2.2 --- Path Switching --- p.14
Chapter 2.2.1 --- Basic Concept of Path Switching --- p.14
Chapter 2.2.2 --- Capacity Allocation and Route Assignment --- p.15
Chapter 3. --- Path Switching over Benes Network --- p.20
Chapter 3.1 --- The Model of path-switched Benes Network --- p.21
Chapter 3.2 --- Module-to-Module Implementation --- p.21
Chapter 3.2.1 --- The First Stage (Input Module) --- p.22
Chapter 3.2.2 --- The Middle Stage (Central Module) --- p.23
Chapter 3.2.3 --- The Last Stage (Output Module) --- p.24
Chapter 3.3 --- Port-to-Port Implementation --- p.24
Chapter 3.3.1 --- Uniform Traffic --- p.25
Chapter 3.3.2 --- Mult irate Traffic --- p.26
Chapter 3.4 --- Closing remarks --- p.29
Chapter 4. --- Performance Analysis --- p.31
Chapter 4.1 --- Traffic Constraints and Perform- ance Guarantees --- p.32
Chapter 4.1.1 --- Arrival Curve and Service Curve --- p.33
Chapter 4.1.2 --- Delay Bound and Backlog Bound --- p.36
Chapter 4.2 --- Service Guarantees --- p.39
Chapter 4.3 --- Deterministic Bounds --- p.42
Chapter 4.3.1 --- Delay --- p.42
Chapter 4.3.2 --- Backlog at Input Module --- p.44
Chapter 4.3.3 --- Backlog at Output Module --- p.47
Chapter 5. --- Simulation Results --- p.52
Chapter 5.1 --- Uniform Traffic --- p.53
Chapter 5.2 --- Multirate Traffic --- p.55
Chapter 6. --- Conclusions and Future Research --- p.59
Chapter 6.1 --- Suggestions for future research --- p.61
Bibliography --- p.62