Made available in DSpace on 2014-12-17T15:48:12Z (GMT). No. of bitstreams: 1 CamilaA.pdf: 551643 bytes, checksum: c211e0d0bbaf86da86337efffe6f407b (MD5) Previous issue date: 2007-07-28 / The Reconfigurables Architectures had appeares as an alternative to the ASICs and the GGP, keeping a balance between flexibility and performance. This work presents a proposal for the modeling of Reconfigurables with Chu Spaces, describing the subjects main about this thematic. The solution proposal consists of a modeling that uses a generalization of the Chu Spaces, called of Chu nets, to model the configurations of a Reconfigurables Architectures. To validate the models, three algorithms had been developed and implemented to compose configurable logic blocks, detection of controllability and observability in applications for Reconfigurables Architectures modeled by Chu nets / As Arquiteturas Reconfigur?veis surgiram no ambiente acad?mico como uma alternativa aos ASICs e aos GGP, mantendo um equil?brio entre flexibilidade e performance. Este trabalho apresenta uma proposta para a modelagem de Arquiteturas Reconfigur?veis com Espa?os de Chu, descrevendo os principais assuntos relativos a esta tem?tica. A solu??o proposta consiste em uma modelagem que utiliza uma generaliza??o dos Espa?os de Chu, denominada de Chu nets, para modelar as configura??es de uma Arquitetura Reconfigur?vel. Como forma de validar os modelos, foram desenvolvidos e implementados tr?s algoritmos que realizam a composi??o de c?lulas l?gicas program?veis, detec??o dos vetores de controlabilidade e observabilidade em aplica??es para Arquiteturas Reconfigur?veis, que est?o modeladas atrav?s das Chu nets

This dissertation investigates high-level decision making for agents that are both goal and utility driven. We develop a partially observable Markov decision process (POMDP) planner which is an extension of an agent programming language called DTGolog, itself an extension of the Golog language. Golog is based on a logic for reasoning about action—the situation calculus. A POMDP planner on its own cannot cope well with dynamically changing environments and complicated goals. This is exactly a strength of the belief-desire-intention (BDI) model: BDI theory has been developed to design agents that can select goals intelligently, dynamically abandon and adopt new goals, and yet commit to intentions for achieving goals. The contribution of this research is twofold: (1) developing a relational POMDP planner for cognitive robotics, (2) specifying a preliminary BDI architecture that can deal with stochasticity in action and perception, by employing the planner. / Computing / M. Sc. (Computer Science)

Submitted by Leonardo Cavalcante (leo.ocavalcante@gmail.com) on 2018-05-03T13:44:12Z No. of bitstreams: 1 Arquivototal.pdf: 1150863 bytes, checksum: a7e25ab87986c9d088c0fe224303f97f (MD5) / Made available in DSpace on 2018-05-03T13:44:12Z (GMT). No. of bitstreams: 1 Arquivototal.pdf: 1150863 bytes, checksum: a7e25ab87986c9d088c0fe224303f97f (MD5) Previous issue date: 2017-03-31 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / In this thesis we presents results on the exact controllability of the partial Di erential Equations (PDEs) of the parabolic and hyperbolic type, in the context of hierarchic control, using the Stackelberg-Nash strategy. In every problems we consider a main control (leader) and two secondary controls (followers). To each leader we obtain a correnponding Nash equilibrium, associated to a bi-objective optimal control problem; then we look for a leader of minimal cost that solves the exact controllability problem. For the parabolic problems we have distributed and boundary controls, now in the hyperbolics every controls are distributed. We consider linear and semilinear cases, which we solve using observability inequality obtained combining right Carleman inequalities. Also we use a xed point method. / Nesta tese apresentamos resultados sobre controlabilidade exata de Equações Diferenciais Parciais (EDPs) dos tipos parabólico e hiperbólico, no contexto de controle hierárquico, usando a estratégia de Stackelberg-Nash. Em todos os problemas consideramos um controle principal (líder) e dois controles secundários (seguidores). Para cada líder obtemos um equil íbrio de Nash correspondente, associado a um problema de controle ótimo bi-objetivo; então buscamos o líder de custo que resolve o problema de controlabilidade. Para os problemas parabólicos temos controles distribuídos e na fronteira, já nos hiperbólico todos os controles são distribuídos. Consideramos casos lineares e semilineares, os quais resolvemos usando desigualdade de observabilidade obtidas combinando desigualdades de Carleman adequadas. Também usamos um método de ponto xo.

Neste trabalho, introduzimos os conceitos de controlabilidade e de observabilidade para equações diferenciais ordinárias generalizadas, apresentamos resultados inéditos sobre condições suficientes e necessárias para controlabilidade e para observabilidade para estas equações e também apresentaremos uma aplicação. Utilizando teoremas de correspondência entre equações diferenciais ordinárias generalizadas e outras equações diferenciais, traduzimos os resultados obtidos para os casos particulares de controlabilidade e observabilidade para equações diferenciais em medida e equações diferencias com impulsos. O fato de trabalharmos no ambiente das equações diferenciais ordinárias generalizadas permitiu que os resultados obtidos pudessem envolver funções com muitas descontinuidades e muito oscilantes, ou seja, de variação ilimitada. Os resultados novos apresentados aqui estão contidos no artigo [21] que se encontra em fase final de redação e será submetido à publicação em breve. / In this work, we introduce concepts of controllability and observability for generalized ordinary differential equations, we present new results on necessary and sufficient conditions for controllability and observability for these equations and we also present an application. Using theorems of correspondence between generalized ordinary differential equations and other differential equations, we translate the results obtained for the particular cases of controllability and observability for measure differential equations and differential equations with impulses. The fact that we work in the framework of generalized ordinary differential equations allows us to obtain results where the functions involved can have many discontinuities and be highly oscillating, that is, of unbounded variation. The new results presented here are contained in the preprint [21] which is under final revision and will soon be submitted for publication.

Dans un objectif conjoint d'identification paramétrique en ligne, les méthodes développées dans cette thèse permettent de concevoir en ligne et en boucle fermée les entrées optimales qui enrichissent les informations contenues dans l'expérience en cours. Ces méthodes reposent sur des mesures en temps réel du procédé, sur un modèle dynamique non linéaire (ou linéaire) multi-variable choisi du procédé, sur un modèle de sensibilité des mesures par rapport aux paramètres à estimer et sur un observateur non linéaire. L'analyse de l'observabilité et des techniques de commande prédictive permettent de définir la commande optimale qui est déterminée en ligne par optimisation sous contraintes. Des aspects de stabilisation sont également étudiés (via un apport de contraintes fictives ou via une technique de Lyapunov). Enfin, une loi de commande explicite pour le cas particulier du système d'ordre un est développée. Des exemples illustratifs sont traités via le logiciel ODOE4OPE : un bioréacteur, un réacteur continu parfaitement agité et une aile delta. Ces exemples permettent de voir que l'estimation des paramètres peut être réalisée avec une bonne précision, et à moindre coût expérimental en une expérience / For online parameter identification, the developed methods here allow to design online and in closed loop optimal inputs that enrich the information in the current experience. These methods are based on real-time measurements of the process, on a dynamic nonlinear (or linear) multi-variable model, on a sensitivity model of measurements with respect to the parameters to be estimated and a nonlinear observer. Analysis of observability and predictive control techniques are used to define the optimal control which is determined online by constrained optimization. Stabilization aspects are also studied (by adding fictitious constraints or by a Lyapunov technique). Finally, for the particular case of a first order linear system, the explicit control law is developed. Illustrative examples are processed via the ODOE4OPE software : a bio-reactor, a continuous stirred tank reactor and a delta wing. These examples help to see that the parameter estimation can be performed with good accuracy in a single and less costly experiment

This thesis deals with state estimation methods for AC drives sensorless control and with possibilities of the estimation. Conditions for observability for a synchronous drive were derived, as well as conditions for the moment of inertia and the load torque observability for both drive types - synchronous and asynchronous. The possibilities of the estimation were confirmed by experimental results. The covariance matrices for all filters were found using an EM algorithm. Both drives were also identified. The algoritms used for state estimation are Extended Kalman Filter, Unscented Kalman Filter, Particle Filters and Moving Horizon Estimator.

Background:                                                                                                            Acquiring comprehension and observability of software systems is a vital and necessary activity for testing and maintenance; however, these tasks are time-consuming for engineers. Concurrently cloud computing requires microservices to enhance the utilization of cloudnative deployment, which simultaneously introduces a high degree of complexity. Further,codifying and distributing technical knowledge within the organization has been proven to be vital for both competitiveness and financial performance. However, doing it successfully has been proven to be difficult, and transitioning to working virtually and in DevOps brings new potential challenges for software firms. Objective:                                                                                                              The objective of this study is to explore how system comprehension of a microservice architecture can be improved from performance metrics through an exploratory data analysis approach. To further enhance the practical business value, the thesis also aims to explore the effects transitioning to virtual work and DevOps have had on knowledge sharing for software firms. Method:                                                                                                                    A case study is conducted at Ericsson with performance data generated from testing of a system deployed in Kubernetes. Data is extracted with Prometheus, and the performance behavior of four interacting pods is explored with correlation analysis and visualization tools.Furthermore, to explore virtual work and DevOps effects on intra-organizational knowledge sharing of technical knowledge, semi-structured interviews were cross analyzed with literature. Results:                                                                                                                  An overall high correlation between performance metrics could be observed with deviations between test cases. Also, we were able to generate propositions regarding the performance behavior as well as bring forward possible candidates for predictive modeling. Four new potential decisive factors driving the choice of activities and transfer mechanisms for knowledge transfer are identified, namely, accessibility, dynamicity, established processes, and efficiency. The transition to virtual work showed five positive factors and three negatives. Effects from DevOps were mostly connected to the frequency of sharing and the potentials of automation.  Conclusions: Our findings suggest that correlation analysis, when used along with visualization tools, can improve system comprehension of cloud-native systems. And while it shows promise for analyzing individual services and hypothesis creation, the method utilized in the study showcased some drawbacks which are covered in the discussion. The findings also point towards the fact that performance metrics can be a rich information source for knowledge and thus deserves further investigation.Findings also suggest that knowledge sharing is not only considered an important element by academia but also deliberately practiced by industry agents. Looking at the transition to virtual work and DevOps, the results imply that they affect knowledge transfer, both in combination and isolation. However, the case study findings do point towards the fact that the transition to working virtually potentially exerts a larger influence. Interviewees expressed both positive and negative aspects of virtual knowledge sharing. Simultaneously, the positive influences of DevOps were followed by extensive challenges. / Bakgrund:  Att erhålla förståelse och observerbarhet av mjukvarusystem är en vital och nödvändig aktivitet, speciellt för testning och underhåll. Samtidigt så är dessa uppgifter både komplexa och tidskrävande för ingenjörer. Mikroservicearkitekturen som utnyttjas för att bygga molnintegrerade lösningar introducerar en hög grad av komplexitet. Fortsättningsvis, att kodifiera och distribuera teknisk kunskap har visats vara kritiskt för organisationers konkurrenskraft och finansiella resultat. Att göra det framgångsrik har dock flertal utmaningar och när flera mjukvarubolag under senare tid övergått till att arbeta virtuellt samt skiftat till DevOps har flertalet nya potentiella utmaningar uppdagats. Syfte:  Målet med denna studie är att utforska hur systemförståelse av mjukvarusystem baserade på en mikroservicearkitektur kan förbättras utifrån prestandamätningar med hjälp av undersökande dataanalysmetoder. För att ytterligare utöka det praktiska affärsvärdet så avser avhandlingen även att undersöka effekterna som övergången till virtuellt arbete och DevOps har haft på denintern kunskapsspridning inom mjukvarubolag.  Metod: En fallstudie utförs på Ericsson AB med prestandadata som genererats under testkörningar av ett system som kör på Kubernetes. Data extraherad med Prometheus och prestationsbeteendet utav fyra interagerande ”pods” utforskas genom korrelationsanalys och visualiseringsverktyg. För att undersöka effekterna från virtuellt arbete samt DevOps har på intraorganisatorisk kunskapsdelning av teknisk kunskap så utförs semi-strukturerade intervjuer som sedan korsanalyseras med litteratur. Resultat:  Överlag så uppvisas hög korrelation mellan prestandamätvärden samtidigt som tydliga avvikelser observerades mellan testfall. Utöver detta så generades propositioner angående prestationsbeteendet samtidigt som potentiella kandidater för prediktiv modellering framhävs. Fyra nya potentiella determinanter identifieras för valet av aktiviteter samt överföringsmekanism, nämligen tillgänglighet, dynamik, etablerade processer, och effektivitet. Övergången till virtuellt arbete uppvisade främst fem positiva faktorer och tre negativa. Effekterna utav DevOps var särskilt kopplade till frekvensen av delning samt potential för automation. Slutsatser: Våra resultat tyder på att korrelationsanalys i kombination med visualiseringsverktyg kan användas för att skapa systemförståelse av molnbaserade system. Samtidigt som metoden visar potential för att analysera individuella tjänster och generera hypoteser så påvisar metoden i vår studie vissa nackdelar vilket tas upp i diskussionen. Resultatet tyder dessutom på att prestandadata kan vara en rik informationskälla för kunskapsskapande och bör vara av intresse för ytterligare studier.Resultaten av den kvalitativa undersökning indikerar att kunskapshantering inte bara är ett viktigt element ur akademins perspektiv men även något som omsorgsfullt praktiseras av industrin. Resultatet angående övergången till virtuellt arbete samt DevOps antyder på att båda har inflytande på hur kunskapsspridning bedrivs, både var för sig och i kombination. Samtidigt pekar våra undersökningsresultat på att övergången till att arbeta virtuellt potentiellt har påverkat kunskapshantering i betydligt större utsträckning än DevOps. Intervjuerna uppvisade både positiva och negativa aspekter utav den virtuella påverkan samtidigt som de positiva effekter som uppmättes av DevOps uppföljdes av omfattande utmaningar.

Les stratégies de commande avancée des machines électriques nécessitent une connaissance précise de la position et/ou de la vitesse du rotor. Ces grandeurs mécaniques sont traditionnellement mesurées par des capteurs, ce qui augmente le coût et diminue la fiabilité et la robustesse du système. D’où l’importance de la commande sans capteurs mécaniques, dite commande sensorless : elle consiste à remplacer les capteurs physiques par un observateur d’état, qui estime les grandeurs mécaniques en se basant sur un modèle de la machine et à partir de ses entrées (tensions) et ses sorties (courants mesurés). Toutefois, avant d’entamer la synthèse d’un observateur, il faut examiner l’observabilité du système, c.à.d. vérifier si les grandeurs à estimer peuvent être reconstruites à partir des mesures et des entrées du système.Ce travail de thèse concerne la modélisation et l’étude d’observabilité des machines électriques en vue de la commande sensorless. Dans un premier temps, les modèles des machines électriques sont établis, et il est montré qu’une modélisation unifiée des machines à courant alternatif est possible. Ensuite, en se basant sur ces modèles, l’observabilité des machines électriques est étudiée en vue de la commande sensorless. La théorie d’observabilité instantanée locale est appliquée, ce qui permet de formuler des conditions analytiques, faciles à interpréter et à vérifier en temps-réel, et qui font le lien avec les phénomènes physiques dans les machines. La validité des conditions d’observabilité est confirmée par des simulations et sur des données expérimentales, en appliquant un observateur de type Kalman étendu.Cette thèse contribue à une nouvelle vision des machines à courant alternatif commandées sans capteurs mécaniques, et à une compréhension plus profonde de leurs propriétés. Elle contribue ainsi à la synthèse de stratégies d’observation plus performantes dans les régimes de fonctionnement critiques (à vitesse nulle et/ou à fréquence d’entrée nulle).Les nouveaux concepts proposés dans cette thèse, tels que le flux équivalent et le vecteur d’observabilité, avec les résultats obtenus, ouvrent de nouveaux horizons dans un domaine qui paraît devenir assez mature. / High-performance control of electric drives requires an accurate knowledge of the rotor position and/or speed. These mechanical variables are traditionally measured using sensors, which increases the cost and reduces both the robustness and the reliability of the system. This emphasizes the importance of electric drives control without shaft sensors, often referred to as sensorless control : it consists of replacing sensors with a state observer algorithm, that estimates the desired mechanical variables from currents and voltages sensing and based on the system’s model. Nevertheless, before designing a state observer, the observability ofthe system should be examined, that is, it should be checked whether the states to be estimated can be reconstructed, unambiguously, from the input/output signalsof the system.This work addresses the modeling and the observability analysis of electric drives in the view of mechanical sensors removal. Firstly, electrical machines models areelaborated, and it is shown that a unified modeling of alternating current machines is feasible, for the purpose of designing unified control and estimation strategies.The observability of the machines’ models is next studied in the view of sensorless control. The local instantaneous observability theory is applied, which enables us to formulate physically insightful analytic conditions that can be easily interpreted and tested in real time. The validity of the observability conditions is confirmed by numerical simulations and experimental data, using an extended Kalman observer.This work contributes to novel outlooks on the sensorless alternating current drives and to a deeper understanding of its properties, in order to develop higher performance estimation techniques in the critical operating conditions (mainly at standstill and/or zero-stator-frequency).The concepts introduced throughout this thesis, such as the equivalent flux and the observability vector, with the obtained results, open new horizons in a domainthat seems to become mature enough.

Les avionneurs expriment le besoin de développement de l’avion plus « électrique ». Cela se traduit par des besoins nouveaux en matière de systèmes de motorisation électrique, en termes de puissance à fournir et de partage de ressources matérielles en vue de minimiser la masse embarquée et les coûts. Parmi les études en cours sur l’évolution des machines tournantes, un intérêt particulier porte sur le développement des alterno-démarreurs de type machines synchrones sans balais et à plusieurs étages (BSSG). Celles-ci sont susceptibles de fournir un fort couple lors des phases de démarrage des réacteurs auxquelles elles sont associées. Pour ce faire, la connaissance, à tout moment, de la position du rotor est essentielle. Cependant, l'ajout d’un capteur dédié impacte la conception de la machine, rajoutant du volume, du câblage et augmentant le coût. La réalisation d'une commande dite « sensorless » permettrait de s'affranchir de l'utilisation d'un tel capteur et de simplifier le design des alterno-démarreurs.A partir d'une modélisation fine de la machine, cette thèse étudie les conditions dans lesquelles une telle commande est réalisable et analyse plusieurs techniques permettant d'y parvenir. Une nouvelle méthode d'estimation de la position du rotor, spécifique aux BSSG est proposée, puis illustrée avec des résultats expérimentaux. Cette technique est basée sur le traitement des composantes harmoniques existantes naturellement au stator de la machine et permet l'estimation de la position à l'arrêt et à très basse vitesse. Afin d'étendre l'estimation sur toute la plage de vitesse, une étude d'estimation de position par un observateur d'état à base du modèle complet de la machine en considérant les harmonique injectés (ou existantes) dans les courants du stator est proposée. Cet observateur peut s’appliquer à la machine synchrone à trois étages mais aussi à toute machine synchrone. Dans cette étude, son fonctionnement est illustré sur une machine synchrone à aimants permanents. / The aircraft manufacturers express the need to develop a more "electric" aircraft. This brings forward new requirements for the electric drive systems in terms of increasing the available on-board power and resource sharing in order to optimize the overall mass and cost. Among the ongoing studies on the evolution of motor drives, a particular interest is given to the development of multi-level brushless synchronous starter/generators (BSSG). These drives are likely to provide the high torque required to start-up the reactors to which they are associated. For this purpose, the knowledge, at any time, of the rotor position is essential. However, adding a dedicated sensor impacts on the design of the machine, increasing volume, cabling needs and cost. For this purpose, investigating on “sensorless” control laws will permit to avoid using such a sensor and to simplify the design of the Starter/Generators.Using a fine modelling of the machine, this work studies the conditions of feasibility for sensorless control and analyzes several techniques for this purpose. A new method of estimation of the shaft-position, particular to the BSSG architecture is proposed and then illustrated with experimental results. This technique is based on the processing of the existing harmonic components naturally in the stator of the machine and allows the estimation of the position at standstill and the very low speed. To extend the estimation to the whole speed range, a study of position estimation using a state observer using the complete model of the machine considering the knowledge of the existing (or injected) harmonic components in the stator currents is proposed. This observer can be applied to the brushless synchronous starter/generator but also on generic synchronous machines. In this study, its performance is illustrated on a permanent magnet synchronous machine.

The increasing demand for electric vehicles (EVs) has led to technological advancementsin the field of battery technology. State of charge (SOC) estimation is a vital function ofthe battery management system - the heart of EVs, and Kalman filtering is a commonmethod for SOC estimation. Due to the non uniformities in tuning and testing scenarios,quantifying performance of SOC estimation algorithms is difficult. Gathering data fordifferent operational scenarios is also cumbersome. In this thesis, SOC estimation algorithmsare developed and tested for a variety of scenarios like varying sensor noise andbias properties, varying state and parameter initializations as well as different initial celltemperatures. A validated and open-source simulation plant model is used to enable easygathering of data for different operational scenarios.The simulation results show that unscented Kalman filter performs better than extendedKalman filter in presence of hard nonlinearities and high initial uncertainties. However,both filters gave similar performance under nominal conditions implying that the choiceof estimation algorithms must depend on operational scenarios. Observability analysisalso gave valuable information to aid in selection of algorithms. The simulation plantmodel facilitated easy data collection for initial development of algorithms, which werethen tested successfully using a real dataset. Further testing using real datasets is requiredto enhance validation. / Den ¨okande efterfr°agan p°a elfordon har lett till teknologiska framsteg inom omr°adet batteriteknik.Estimering av batteriets laddningstillst°and ¨ar en essentiell funktion i batteristyrsystemet,hj¨artat i ett elfordon, och g¨ors ofta genom att till¨ampa metoden Kalmanfiltrering.P°a grund av varierande implementations och testmetodik i litteraturen ¨ar detsv°art att kvantifiera estimeringsalgoritmer. I denna avhandling utvecklas algoritmer f¨oratt estimera ett batteris laddningstillst°and. Algoritmerna testas f¨or olika former av sensorfeloch initialtillst°and, samt f¨or en rad olika temperaturer. En validerad datormodell avbatteri, sensorer och omgivning nyttjas f¨or att generera representativa data f¨or de olikaf¨orh°allandena.Simuleringsresultat visar att den s°a kallade doftl¨osa varianten av Kalmanfiltret (UKF)presterade b¨attre ¨an det utvidgade Kalmanfiltret (EKF) i fall d¨ar systembeteendet ¨ar mycketolinj¨art och d°a initialtillst°andet ¨ar os¨akert. Under normala f¨orh°allanden presterardock de b°ada algoritmerna likv¨ardigt, vilket antyder att valet av algoritm b¨or g¨oras medavseende anv¨andningsscenario. En observerbarhetsanalys av de olika filtervarianterna gavytterligare v¨ardefull information f¨or valet av algoritm. Efter utveckling av filtreringsalgoritmernai simuleringsmilj¨o utf¨ordes tester p°a faktiska m¨atdata med goda resultat. F¨or attfullst¨andig validering av algoritmerna kr¨avs emellertid mer utt¨ommande tester.