Master thesis is focused on the philosophy of design adaptive controller. In the theoretic part are described parts of the adaptive controller, which belongs parts as online identification by recursive least-squares method and PSD controller, which can set its parameters according to identified system (use Z-N method). The part of control system with transport delay is situated at the conclusion of the theoretic part, there are focused on Smith predictor. Practical part is focused on verification of all algorithms, which was performed on models and real systems.

Les gyroscopes laser à gaz constituent une solution technique de haute performances dans les problématiques de navigation inertielle. Néanmoins, pour de très faibles vitesses de rotation, les petites imperfections des miroirs de la cavité optique font que les deux faisceaux contra-propageant sont verrouillés en phase. En conséquence, les mesures en quadrature de leur différence de phase ne permettent plus de remonter directement aux vitesses de rotation à l'intérieur d'une zone autour de zéro, dite zone aveugle statique, ou, si l'on utilise une procédure d'activation mécanique, dite zone aveugle dynamique. Ce travail montre qu'il est néanmoins possible, en utilisant des méthodes issues du filtrage et de l'estimation, de remonter aux vitesses de rotation mêmes si ces dernières sont en zone aveugle. Pour cela, on part d'une modélisation physique de la dynamique que l'on simplifie par des techniques de perturbations singulières pour en déduire une généralisation des équations de Lamb. Il s'agit de quatre équations différentielles non-linéaires qui décrivent la dynamique des intensités et des phases des deux faisceaux contra-propageant. Une étude qualitative par perturbations régulières, stabilité exponentielle des points d'équilibre et applications de Poincaré permet de caractériser les zones aveugles statiques et dynamiques en fonction des imperfections dues aux miroirs. Il est alors possible d'estimer en ligne avec un observateur asymptotique fondé sur les moindre carrés récursifs ces imperfections en rajoutant aux deux mesures en quadrature celles des deux intensités. La connaissance précise de ces imperfections permet alors de les compenser dans la dynamique de la phase relative, et ainsi d'estimer les rotations en zone aveugle. Des simulations numériques détaillées illustrent l'intérêt de ces observateurs pour augmenter la précision des gyroscopes à gaz. / Gaz ring laser gyroscopes provide a high performance technical solution for inertial navigation. However, for very low rotational speeds, the mirrors imperfections of the optical cavity induce a locking phenomena between the phases of the two counter-propagating Laser beams. Hence, the measurements of the phase difference can no longer be used when the speed is within an area around zero, called lock-in zone, or,if a procedure of mechanical dithering is implemented, dithering lock-in zone. Nevertheless, this work shows that it is possible using filtering and estimation methods to measure the speed even within the lock-in zones. To achieve this result, we exploit a physical modeling of the dynamics that we simplify, using singular perturbation techniques, to obtain a generalization of Lamb's equations. There are four non-linear differential equations describing the dynamics of the intensities and phases of the two counter-propagating beams. A qualitative study by regular perturbation theory, exponential stability of the equilibrium points and Poincaré maps allows a characterisation of the lock-in zones according to the mirrors imperfections. It is then possible to estimate online, with an asymptotic observer based on recursive least squares, these imperfections by considering the additional measurements of the beam intensities. Accurate knowledge of these imperfections enables us to compensate them in the dynamic of the relative phase, and thus to estimate rotational speeds within the lock-in zones. Detailed numerical simulations illustrate the interest of those observers to increase the accuracy of gas ring laser gyroscopes.

Conselho Nacional de Desenvolvimento CientÃfico e TecnolÃgico / Este trabalho trata da proposiÃÃo de algoritmos para equalizaÃÃo cega de canais lineares e nÃao-lineares inspirados no Algoritmo do MÃdulo Constante (CMA). O CMA funciona de maneira bastante eficiente com constelaÃÃes nas quais todos os pontos possuem a mesma amplitude, como em modulaÃÃes do tipo Phase Shift Keying (PSK). Entretanto, quando os pontos da constelaÃÃo podem assumir diferentes valores de amplitudes, como em modulaÃÃes do tipo Quadrature Amplitude Modulation (QAM), o CMA e seus derivados muitas vezes nÃo funcionam de forma satisfatÃria. Desta forma, as tÃcnicas aqui propostas sÃo projetadas para melhorar a performance do CMA em termos de velocidade de convergÃncia e precisÃo, quando operando em sinais transmitidos com diversos mÃdulos, em particular para a modulaÃÃo QAM. Assim como o CMA, para possuir um bom apelo prÃtico, essas tÃcnicas devem apresentar bom compromisso entre complexidade, robustez e desempenho. Para tanto, as tÃcnicas propostas utilizam o Ãltimo sÃmbolo decidido para definir uma estimaÃÃo de raio de referÃncia para a saÃda do equalizador. De fato, esses algoritmos podem ser vistos como generalizaÃÃes do CMA e de alguns derivados do CMA para constelaÃÃes com mÃltiplos raios. A proposiÃÃo de algoritmos do tipo gradiente estocÃstico à concluÃda com o desenvolvimento de tÃcnicas originais, baseadas no CMA, para equalizaÃÃo de canais do tipo Wiener, que consiste em um filtro linear com memÃria, seguido por um filtro nÃo-linear sem memÃria. As expressÃes para a adaptaÃÃo do equalizador sÃo encontradas com o auxÃlio de uma notaÃÃo unificada para trÃs diferentes estruturas: i) um filtro de Hammerstein; ii) um filtro de Volterra diagonal; e iii) um filtro de Volterra completo. Um estudo teÃrico acerca do comportamento do principal algoritmo proposto, o Decision Directed Modulus Algorithm (DDMA) à realizado. SÃo analisadas a convergÃncia e a estabilidade do algoritmo atravÃs de uma anÃlise dos pontos de mÃnimo de sua funÃÃo custo. Outro objetivo à encontrar o valor teÃrico do Erro MÃdio QuadrÃtico MÃdio em Excesso - Excess Mean Square Error (EMSE) fornecido pelo DDMA considerando-se o caso sem ruÃdo. Ao final, à feito um estudo em que se constata que o algoritmo DDMA possui fortes ligaÃÃes com a soluÃÃo de Wiener e com o CMA. VersÃes normalizadas, bem como versÃes do tipo Recursive Least Squares (RLS), dos algoritmos do tipo gradiente estocÃstico estudados sÃo tambÃm desenvolvidas. Cada famÃlia de algoritmos estudada fie composta por quatro algoritmos com algumas propriedades interessantes e vantagens sobre as tÃcnicas clÃssicas, especialmente quando operando em sinais QAM de ordem elevada. TambÃm sÃo desenvolvidas versÃes normalizadas e do tipo RLS dos algoritmos do tipo CMA estudados para equalizaÃÃo de canais nÃo-lineares. O comportamento de todas as famÃlias de algoritmos desenvolvidos à testado atravÃs de simulaÃÃes computacionais, em que à verificado que as tÃcnicas propostas fornecem ganhos significativos em desempenho, em termos de velocidade de convergÃncia e erro residual, em relaÃÃo Ãs tÃcnicas clÃssicas. / This work studies and proposes algorithms to perform blind equalization of linear and nonlinear channels inspired on the Constant Modulus Algorithm (CMA). The CMA works very well for modulations in which all points of the signal constellation have the same radius, like in Phase Shift Keying (PSK) modulations. However, when the constellation points are characterized by multiple radii, like in Quadrature Amplitude Modulation (QAM) signals, the CMA does not work properly in many situations. Thus, the techniques proposed here are designed to improve the performance of the CMA, in terms of speed of convergence and residual error, when working with signals transmitted with multiple magnitude, in particular with QAM signals. As well as for the CMA, these techniques should have a good compromise among performance, complexity and robustness. To do so, the techniques use the last decided symbol to estimate reference radius to the output of the equalizer. In fact, they can be seen as modifications of the CMA and of some of its derivatives for constellations with multiple radii. The proposition of stochastic gradient algorithms is concluded with the development of new adaptive blind techniques to equalize channels with a Wiener structure. A Wiener filter consists of a linear block with memory followed by a memoryless nonlinearity, by using the CMA. We develop expressions for the adaptation of the equalizer using a unified notation for three different equalizer filter structures: i) a Hammerstein filter, ii) a diagonal Volterra filter and iii) a Volterra filter. A theoretical analysis of the main proposed technique, the Decision Directed Modulus Algorithm (DDMA), is also done. We study the convergence and the stability of the DDMA by means of an analysis of the minima of the DDM cost function. We also develop an analytic expression for the Excess Mean Square Error (EMSE) provided by the DDMA in the noiseless case. Then, we nd some interesting relationships among the DDM, the CM and the Wiener cost functions. We also develop a class of normalized algorithms and a class of Recursive Least Squares (RLS)-type algorithms for blind equalization inspired on the CMA-based techniques studied. Each family is composed of four algorithms with desirable properties and advantages over the original CM algorithms, specially when working with high-level QAM signals. Normalized and RLS techniques for equalization of Wiener channels are also developed. The behavior of the proposed classes of algorithms discussed is tested by computational simulations. We verify that the proposed techniques provide significative gains in performance, in terms of speed of convergence and residual error, when compared to the classical algorithms.

Made available in DSpace on 2016-12-12T17:38:34Z (GMT). No. of bitstreams: 1 Paulo Cesar da Silva.pdf: 5331875 bytes, checksum: adf5bf8880e8e89421a2fd553c966530 (MD5) Previous issue date: 2015-07-10 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The electrical parameters of three-phase induction motors vary according to their operating point, with the temperature and also with the natural machine degradation. Since the design regulators are typically performed based on linear and simplified equations of the machine, the parametric variation can cause unwanted responses, since the motor behavior is non-linear and time variant. Thus, it is proposed in this work carry the identification of the electrical parameters of the induction motor using the estimator called Least Squares Recursive (RLS). Thus, with the parametric data updated every sampling period can be recalculated in real time the gains of the regulators are designed for controlling the induction machine and make minor mismatches. The experimental results were obtained with the processing performed by dSPACE hardware (DS1103), which has an interface with Matlab/Simulink, facilitating the use by the user and reducing the time taken for testing bench. The results of numerical and practical simulations show a comparison between the mesh proposal, the parametric identification and update of the gains of the controllers (adaptive control) and the mesh with controllers with fixed gains. / Os parâmetros elétricos do motor de indução trifásico variam conforme o seu ponto de operação, com a temperatura e também com a degradação natural da máquina. Visto que o projeto de reguladores são tipicamente realizados com base em equações lineares e simplicadas da máquina, a variação paramétrica pode causar respostas indesejadas, pois o comportamento do motor é não-linear e variante no tempo. Desta forma propõe-se neste trabalho realizar a identificação dos parâmetros elétricos do motor de indução, utilizando o estimador denominado de Mínimos Quadrados Recursivos (MQR). Assim, com os dados paramétricos atualizados a cada período de amostragem, pode-se recalcular em tempo real, os ganhos dos reguladores que são projetados para o controle da máquina de indução e tornar os descasamentos menores. Os resultados experimentais foram obtidos com o processamento realizado pelo hardware dSPACE (DS1103), que possui uma interface com o Matlab/Simulink, facilitando a utilização por parte do usuário e reduzindo o tempo dispendido para os testes em bancada. Os resultados de simulações numéricas e práticos apresentam uma comparação entre a malha proposta, com a identificação paramétrica e atualiza ção dos ganhos dos controladores (controle adaptativo) e a malha com controladores com ganhos fixos.

Made available in DSpace on 2014-12-17T14:55:56Z (GMT). No. of bitstreams: 1 EllonPM_DISSERT.pdf: 1231242 bytes, checksum: 49456bef5c0d0bfdc5bf49d689568b60 (MD5) Previous issue date: 2012-01-27 / Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico / Several mobile robots show non-linear behavior, mainly due friction phenomena between the mechanical parts of the robot or between the robot and the ground. Linear models are efficient in some cases, but it is necessary take the robot non-linearity in consideration when precise displacement and positioning are desired. In this work a parametric model identification procedure for a mobile robot with differential drive that considers the dead-zone in the robot actuators is proposed. The method consists in dividing the system into Hammerstein systems and then uses the key-term separation principle to present the input-output relations which shows the parameters from both linear and non-linear blocks. The parameters are then simultaneously estimated through a recursive least squares algorithm. The results shows that is possible to identify the dead-zone thresholds together with the linear parameters / V?rios rob?s m?veis apresentam comportamentos n?o-lineares, principalmente ocasionados por fen?menos de atrito entre as partes mec?nicas do rob? ou entre o rob? e o solo. Modelagens puramente lineares apresentam-se eficientes em alguns casos, mas ? preciso levar em considera??o as n?o-linearidades do rob? quando se deseja movimentos ou posicionamentos precisos. Este trabalho prop?e um procedimento de identifica??o param?trica do modelo de um rob? m?vel com acionamento diferencial, no qual s?o consideradas as n?o-linearidades do tipo zona-morta presentes nos atuadores do rob?. A proposta baseia-se no modelo de Hammerstein para dividir o sistema em blocos lineares e n?o-lineares. O princ?pio da separa??o do termo chave ? utilizado para demonstrar a rela??o entre as entradas e sa?das do sistema com os par?metros tanto da parcela linear quanto da n?o-linear. Os par?metros de ambas as parcelas s?o identificados simultaneamente, atrav?s de um algoritmo de m?nimos quadrados recursivo. Os resultados mostram que ? poss?vel identificar o valor os limites da zona-morta assim como os par?metros da parcela linear do modelo do sistema

Made available in DSpace on 2016-08-17T14:53:22Z (GMT). No. of bitstreams: 1 Dissertacao Allan James.pdf: 3170694 bytes, checksum: 054a9e74e81a7c2099800246d0b6c530 (MD5) Previous issue date: 2012-09-28 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The union of methodologies for optimal control and dynamics programming has stimulated the development of algorithms for realization of discrete control systems of the type linear quadratic regulator (DLQR). The methodology is based on reinforcement learning methods based on temporal differences and approximate dynamic programming. The proposed method combines the approach of the value function by method RLS (recursive least squares) and approximate policy iteration schemes heuristic dynamic programming (HDP). The approach is directed to the assessment of convergence of the solution DLQR and the heuristic weighting matrices 􀜳 and 􀜴 of the utility function associated with DLQR. The investigation of convergence properties related to consistency, persistent excitation and polarization of the RLS estimator is performed. The methodology involved in a project achievements online DLQR controllers and is evaluated in a fourth order multivariable dynamic system. / A união das metodologias de controle ótimo e de programação dinâmica tem impulsionado o desenvolvimento de algoritmos para realizações de sistemas de controle discreto do tipo regulador linear quadrático (DLQR). A metodologia utilizada neste trabalho é fundamentada sobre métodos de aprendizagem por reforço baseados em diferenças temporais e programação dinâmica aproximada. O método proposto combina a aproximação da função valor através do método RLS (mínimos quadrados recursivos) e iteração de política aproximada em esquemas de programação dinâmica heurística (HDP). A abordagem é orientada para a avaliação da convergência da solução DLQR e para a sintonia heurística das matrizes de ponderação 􀜳 e 􀜴da função de utilidade associada ao DLQR. É realizada a investigação das propriedades de convergência relacionadas à consistência, excitação persistente e polarização do estimador RLS. A metodologia contempla realizações de projetos de forma online de controladores DLQR e é avaliada em um sistema dinâmico multivariável de quarta ordem.

This thesis deals with implementation of industrial used controlling block into Power Panel equipment of B&R in integrated development environment Automation Studio 3.0.71. It is PID controller with autotuner and visualization, which allowed bumpless transfer among controlling algorithms and manual control. PID and I-PD controllers with filtering of derivative action and dynamic antiwindup were implemented. Parameters of PID controller is possible to tune with the aid of ultimate gain and ultimate period according rules of Ziegler-Nichols. Ultimate parameters of controlled plant is possible to acquire with the aid of two identification methods, recursive least squares method and relay feedback. Recursive least squares method was implemented with directional forgetting. For relay feedback were used two types of relays: ideal and saturation relay for improving accuracy of searched ultimate values. The whole solution is programmed in ANSI C which Automation Studio supports. Created controller is control with the aid of touchscreen which is integrated in Power Panel. Trends of process values are viewed on the screen too. For comparison adaptive controller by B&R was implemented. This contoller is standardly supplied with Automation Studio in LoopConR library. All created algorithms were first validated on mathematical model of plant and then on real model in laboratory. The first part of thesis deals with theoretic analysis of used methods. The practical realization is described in the second part of this diploma thesis.

This final thesis deals with adaptive control of small brewery. Contains a short description of brewing technology and adaptive control. Includes design of small brawery and control system based on STM32F407 Cortex-M4 microcontroller. For online identification is used Recursive Least Squares Method. The controller function perform a modified PSD controller with filtered derivative component, reducing the first overshot and dynamic antiwindup. The proposed adaptive controller is implemented into microcontroller. Created device i used for semi-automatic beer production.

Real-Time Rollover prevention of Unmanned Ground Vehicle (UGV) is very paramount to its reliability and survivability mostly when operating on unknown and rough terrains like mines or other planets.Therefore this research presents the method of real-time rollover prevention of UGVs making use of Adaptive control techniques based on Recursive least Squares (RLS) estimation of unknown parameters, in order to enable the UGVs to adapt to unknown hush terrains thereby increasing their reliability and survivability. The adaptation is achieved by using indirect adaptive control technique where the controller parameters are computed in real time based on the online estimation of the plant’s (UGV) parameters (Rollover index and Roll Angle) and desired UGV’s performance in order to appropriately adjust the UGV speed and suspension actuators to counter-act the vehicle rollover. A great challenge of indirect adaptive control system is online parameter identification, where in this case the RLS based estimator is used to estimate the vehicles rollover index and Roll Angle from lateral acceleration measurements and height of the centre of gravity of the UGV. RLS is suitable for online parameter identification due to its nature of updating parameter estimate at each sample time. The performance of the adaptive control algorithms and techniques is evaluated using Matlab Simulink® system model with the UGV Model built using SimMechanics physical modelling platform and the whole system runs within Simulink environment to emulate real world application. The simulation results of the proposed adaptive control algorithm based on RLS estimation, show that the adaptive control algorithm does prevent or minimize the likely hood of vehicle rollover in real time. / Electrical and Mining Engineering / M. Tech. (Electrical Engineering)

No description available.