This thesis addresses the pose stabilization of a car-like vehicle using omnidirectional visual feedback. The presented method allows a vehicle to servo to a pre-learnt target pose based on feature bearing angle and range discrepancies between the vehicle's current view of the environment and that seen at the learnt location. The best example of such a task is the use of visual feedback for autonomous parallel-parking of an automobile. Much of the existing work in pose stabilization is highly theoretical in nature with few examples of implementations on 'real' vehicles, let alone vehicles representative of those found in industry. The work in this thesis develops a suitable test platform and implements vision-based pose stabilization techniques. Many of the existing techniques were found to fail due to vehicle steering and velocity loop dynamics, and more significantly, with steering input saturation. A technique which does cope with the characteristics of 'real' vehicles is to divide the task into predefined stages, essentially dividing the state space into sub-manifolds. For a car-like vehicle, the strategy used is to stabilize the vehicle to the line which has the correct orientation and contains the target location. Once on the line, the vehicle then servos to the desired pose. This strategy can accommodate velocity and steering loop dynamics, and input saturation. It can also allow the use of linear control techniques for system analysis and tuning of control gains. To perform pose stabilization, good estimates of vehicle pose are required. A simple, yet robust, method derived from the visual homing literature is to sum the range vectors to all the landmarks in the workspace and divide by the total number of landmarks--the Improved Average Landmark Vector. By subtracting the IALV at the target location from the currently calculated IALV, an estimate of vehicle pose is obtained. In this work, views of the world are provided by an omnidirectional camera, while a magnetic compass provides a reference direction. The landmarks used are red road cones which are segmented from the omnidirectional colour images using a pre-learnt, two-dimensional lookup table of their colour profile. Range to each landmark is estimated using a model of the optics of the system, based on a flat-Earth assumption. A linked-list based method is used to filter the landmarks over time. Complementary filtering techniques, which combine the vision data with vehicle odometry, are used to improve the quality of the measurements.

This work presents an investigation into the parameter estimation of suspension components and the vertical motions of wheeled vehicles from experimental data. The estimation problems considered were for suspension dampers, a single wheel station and a full vehicle. Using conventional methods (gradient-based (GB), Downhill Simplex (DS)) and stochastic methods (Genetic Algorithm (GA) and Differential Evolution (DE)), three major problems were encountered. These were concerned with the ability and consistency of finding the global optimum solution, time consumption in the estimation process, and the difficulties in setting the algorithm's control parameters. To overcome these problems, a new technique named the discrete variable Hybrid Differential Evolution (dvHDE) method is presented. The new dvHDE method employs an integer-encoding technique and treats all parameters involved in the same unified way as discrete variables, and embeds two mechanisms that can be used to deal with convergence difficulties and reduce the time consumed in the optimisation process. The dvHDE algorithm has been validated against the conventional GB, DS and DE techniques and was shown to be more efficient and effective in all but the simplest cases. Its robustness was demonstrated by its application to a number of vehicle related problems of increasing complexity. These include case studies involving parameter estimation using experimental data from tests on automotive dampers, a single wheel station and a full vehicle. The investigation has shown that the proposed dvHDE method, when compared to the other methods, was the best for finding the global optimum solutions in a short time. It is recommended for nonlinear vehicle suspension models and other similar systems.

Este trabalho aborda diversas técnicas de controle, com o intuito de comparação do desempenho e robustez ao aplicá-los a um conversor Buck. Iniciando pelo controlador PID (Proporcional, Integral e Derivativo), amplamente explorado e dominado no meio industrial, ele é adotado neste trabalho como referência de comparação para as demais técnicas desenvolvidas. Outra estratégia aqui apresentada é o GANLPID (Gaussian Adaptative Non Linear PID ou PID Adaptativo Não Linear Gaussiano), trata-se de uma técnica não linear, possui ganhos variantes em função do erro baseados em uma função gaussiana. O controle por alocação de polos é uma técnica de controle que em sua forma básica não possui parcela integral, sendo necessária a inclusão deste termo para minimizar o erro em regime permanente. As principais características de análise de desempenho são o tempo de acomodação e overshoot. Todas as técnicas são exploradas a fim de serem implementadas em dispositivos FPGA (Field Programmable Gate Array), possuindo algumas vantagens sobre microcontroladores e DSP’s (Digital Signal Processor), pois conseguem executar tarefas em paralelo deixando a execução do algoritmo mais rápida. As técnicas de controle escolhidas foram simuladas utilizando a ferramenta DSP Builder e compiladas diretamente em código HDL (linguagem de descrição de hardware). Os resultados de simulação e experimentais são apresentados e comentados para validar os projetos propostos. / This work discuss several techniques of control, with an intention of comparison of performance and robustness to apply them to Buck coverter. Starting with PID (Proportional, Integral, Derivative) controller, widely explored and dominated in an industrial environment, it’s used in this work as comparison reference for the others techniques developed. Another strategy presented here is the GANLPID (Gaussian Adaptative Non LinearPID), it’s a case of non linear technique, has won variants in function of the based on a Gaussian error function. variants have gains on function of error based on a Gaussian function. The pole placement control technique not having full part in their basic forms, being necessary to include this term to eliminate the steady-state error. The main performance analysis features are the settling time and overshoot. All the techniques are explored in order to be implemented in FPGA (Field Programmable Gate Array) devices, having some advantages over microcontrollers and DSP’s (Digital Signal Processor), because can execute tasks in parallel allowing the implementation of the algorithm more faster. The chosen control techniques were simulated using the DSP Builder tool and and compiled directly in HDL (hardware description language) code. The results of simulation and experimental are presented and discussed in order to validate the proposed projects.

Made available in DSpace on 2014-12-17T14:55:11Z (GMT). No. of bitstreams: 1 CarlosAGF_TESE.pdf: 1739972 bytes, checksum: 7401db4e68ede642dc9d65e00bd935e6 (MD5) Previous issue date: 2012-12-21 / In this work a modification on ANFIS (Adaptive Network Based Fuzzy Inference System) structure is proposed to find a systematic method for nonlinear plants, with large operational range, identification and control, using linear local systems: models and controllers. This method is based on multiple model approach. This way, linear local models are obtained and then those models are combined by the proposed neurofuzzy structure. A metric that allows a satisfactory combination of those models is obtained after the structure training. It results on plant s global identification. A controller is projected for each local model. The global control is obtained by mixing local controllers signals. This is done by the modified ANFIS. The modification on ANFIS architecture allows the two neurofuzzy structures knowledge sharing. So the same metric obtained to combine models can be used to combine controllers. Two cases study are used to validate the new ANFIS structure. The knowledge sharing is evaluated in the second case study. It shows that just one modified ANFIS structure is necessary to combine linear models to identify, a nonlinear plant, and combine linear controllers to control this plant. The proposed method allows the usage of any identification and control techniques for local models and local controllers obtaining. It also reduces the complexity of ANFIS usage for identification and control. This work has prioritized simpler techniques for the identification and control systems to simplify the use of the method / Neste trabalho prop?e-se uma modifica??o na estrutura neurofuzzy ANFIS (Adaptive Network Based Fuzzy Inference System) para a obten??o de um m?todo sistem?tico para identifica??o e controle de plantas com ampla faixa de opera??o e n?o linearidade acentuada, a partir de t?cnicas lineares de identifica??o e controle. Este m?todo se baseia na metodologia de m?ltiplos modelos. Dessa forma, obt?m-se modelos lineares locais e esses s?o combinados pela estrutura neurofuzzy proposta. Uma m?trica que permite combinar adequadamente esses modelos ? obtida ap?s o treinamento dessa estrutura, resultando na identifica??o global da planta. Para cada um desses modelos ? projetado um controlador. O controle global ? obtido a partir da combina??o dos sinais dos controladores locais. Essa mistura ? feita pelo ANFIS modificado. A modifica??o na arquitetura do ANFIS permite o compartilhamento do conhecimento adquirido pelo treinamento da estrutura empregada na combina??o de modelos locais. Assim n?o se faz necess?rio o treinamento da estrutura empregada na mistura de controladores. Avaliaram-se as estruturas modificadas atrav?s de dois estudos de caso. Verificou-se que ? poss?vel treinar apenas um ANFIS, para a obten??o de uma m?trica que permita a combina??o adequada dos modelos lineares, v?lidos localmente, e essa estrutura, j? ajustada, pode ser aplicada na combina??o de controladores lineares, projetados para cada um dos modelos, resultando em um sistema de controle que satisfaz as especifica??es de desempenho previamente estabelecidas. O m?todo proposto possibilita a utiliza??o de quaisquer t?cnicas de identifica??o e controle para a obten??o dos modelos e controladores locais, e a redu??o da complexidade de utiliza??o do ANFIS para identifica??o e controle. Neste trabalho priorizaram-se as t?cnicas mais simples de identifica??o e controle de sistemas de forma a simplificar a utiliza??o do m?todo

O presente trabalho tem como objetivo explorar o uso de técnicas de controle avançados na indústria siderúrgica. Propõe-se uma estratégia de controle do nível do aço da piscina formada entre os rolos de um sistema lingotamento contínuo de tiras (LCT) utilizando a tecnologia twin-roll (rolos duplos). O processo LCT rolos duplos tem por finalidade a produção de tiras solidificadas de espessura constante sob uma força de separação entre os rolos também constante. O nível de aço bem como a força de separação são as variáveis mais críticas para a produção de tiras de aço de alta qualidade. O nível pode ser controlado usando a entrada de aço ou a velocidade de laminação. Entretanto, a velocidade de laminação é usualmente utilizada para regular a força de separação entre os rolos. A estratégia de controle proposta inclui a incorporação de um tundish intermediário submerso na piscina. O controle do nível é então feito a partir da saída de aço do tundish intermediário. Consideramos as técnicas de controle linearizante por realimentação de estado e de controle fuzzy usando ambos os modelos Takagi-Sugeno (T-S) e Mamdani. Resultados de simulação são apresentados para uma planta instalada no Instituto de Pesquisa Tecnológica (IPT) de São Paulo, divisão de metalurgia (DIMET). / The aim of this work is to explore the use of advanced control techniques in the metallurgical industry. A control strategy to regulate the molten steellevel of a strip-casting process is proposed. The process produces a solidified strip of constant thickness given by the roll gap under a constant roll separation force. Along with the molten steel level the rool separation force are the most criticaI process variables. The molten steel level may be controlled using the tundish output flow or the casting speed. However, the casting speed is usually used to control the roll force separation. In the control strategy proposed it is incorporated an intermediary tundish submerse into the pool between the rotating rolls to improve the strip thickness uniformity. The molten steel level is thus controlled by the intermediary tundish output flow. Conventional PI, feedback linearizing plus a fuzzy control term and a fuzzy controller in a cascade configuration are considered. Simulation results are presented considering the real system parameters of a plant installed at the Instituto de Pesquisa Tecnológica (IPT) de São Paulo, Divisão de Metalurgia (DIMET).

Orientador: Edson Bim / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação / Made available in DSpace on 2018-08-16T22:20:50Z (GMT). No. of bitstreams: 1 Franca_AlexPereira_M.pdf: 3058764 bytes, checksum: beb5e6bdbba91505ff0ef14262e6be5c (MD5) Previous issue date: 2010 / Resumo: Uma abordagem preditiva global-local para o controle de um motor de indução é apresentado nesta tese. O conceito de controle preditivo diz respeito a uma classe de controladores que tem se desenvolvido muito no âmbito do controle de sistemas de conversão de energia nos últimos anos, acompanhando o desenvolvimento da capacidade computacional dos sistemas microprocessados ao longo da última década. A técnica proposta é fundamentada numa lei de controle baseada em uin modelo identificado de um motor de indução. A identificação se dá de maneira experimental a partir da simulação de um motor de indução indiretamente orientado em função do fluxo do rotor, através de um modelo fuzzy do tipo Takagi-Sugeno TS com Funções de Base Ortonormal no consequente das regras.As ações de controle locais são combinadas são combinadas pela ativação das regras do modelo local devido, e a ação de controle global resultante é aplicada ao controle de velocidade do motor de indução. Este método permite ao controlador a inclusão nos parâmetros de controle das não-linearidades e restrições inerentes ao controle do máquinas elétricas / Abstract: A predictive global-local approach technic for induction motor control is presented in this thesis. Predictive control is a very wide class of controllers that have found rather recent applications in the control of electrical machines. Research on this topic has been increased in the last years due to the possibilities of today s microprocessors used for control. The proposed technique is founded on a identified model based predictive control. The identification technique applied in this method is based on the black box modeling of a Indirect Rotor Flux Oriented Induction Motor in stationary reference system through Takagi-Sugeno-TS fuzzy models with orthonormal base functions - OBF - on the rules consequents. Control actions are matched by local activation of the model rule because local and global control action is applied to speed control of an induction motor. This method allows the inclusion in the control parameters of non-linearities and constraints inherent the control of electrical machines / Mestrado / Energia Eletrica / Mestre em Engenharia Elétrica

Submitted by Luciana Ferreira (lucgeral@gmail.com) on 2017-12-18T11:49:36Z No. of bitstreams: 2 Dissertação - Douglas Freire de Carvalho - 2017.pdf: 23722781 bytes, checksum: 3e99befa0116cb9ca768d092c541c055 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2017-12-18T11:50:09Z (GMT) No. of bitstreams: 2 Dissertação - Douglas Freire de Carvalho - 2017.pdf: 23722781 bytes, checksum: 3e99befa0116cb9ca768d092c541c055 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-12-18T11:50:09Z (GMT). No. of bitstreams: 2 Dissertação - Douglas Freire de Carvalho - 2017.pdf: 23722781 bytes, checksum: 3e99befa0116cb9ca768d092c541c055 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-11-20 / Conselho Nacional de Pesquisa e Desenvolvimento Científico e Tecnológico - CNPq / This work presents study of nonlinear predictive controller, tuned by optimization process, applied to independent excitation DC motor speed control, controlled by fully controlled three phase rectifier. The methodology starts from the construction of the workbench that represents the real system followed by the development and validation of the computational model. The optimization process promotes the obtaining of optimized parameters for the nonlinear model based predictive controller to be implemented in the real process. Comparisons are made between the control technique proposed with classical control techniques, in computational and real environment. The results indicate the prominence of the proposed technique to be used in control of non-linear systems where seeks smaller errors even with disturbance insertions. / Este trabalho apresenta estudo de controlador preditivo não linear, sintonizado por processo de otimização, aplicado ao controle de velocidade de motor de corrente contínua de excitação independente acionado por retificador trifásico totalmente controlado. A metodologia parte da construção da bancada que representa o sistema real seguida pelo desenvolvimento e validação de modelo computacional. O processo de otimização, promove a obtenção de parâmetros otimizados para o controlador preditivo não linear baseado em modelos a serem implementados no processo real. São realizadas comparações entre a técnica de controle proposta com técnicas de controle clássicas, em ambiente computacional e real. Os resultados indicam a proeminência da técnica proposta a ser empregada em controle de sistemas não lineares onde busca-se menores erros mesmo com inserção de distúrbios.

This work presents an investigation into the parameter estimation of suspension components and the vertical motions of wheeled vehicles from experimental data. The estimation problems considered were for suspension dampers, a single wheel station and a full vehicle. Using conventional methods (gradient-based (GB), Downhill Simplex (DS)) and stochastic methods (Genetic Algorithm (GA) and Differential Evolution (DE)), three major problems were encountered. These were concerned with the ability and consistency of finding the global optimum solution, time consumption in the estimation process, and the difficulties in setting the algorithm's control parameters. To overcome these problems, a new technique named the discrete variable Hybrid Differential Evolution (dvHDE) method is presented. The new dvHDE method employs an integer-encoding technique and treats all parameters involved in the same unified way as discrete variables, and embeds two mechanisms that can be used to deal with convergence difficulties and reduce the time consumed in the optimisation process. The dvHDE algorithm has been validated against the conventional GB, DS and DE techniques and was shown to be more efficient and effective in all but the simplest cases. Its robustness was demonstrated by its application to a number of vehicle related problems of increasing complexity. These include case studies involving parameter estimation using experimental data from tests on automotive dampers, a single wheel station and a full vehicle. The investigation has shown that the proposed dvHDE method, when compared to the other methods, was the best for finding the global optimum solutions in a short time. It is recommended for nonlinear vehicle suspension models and other similar systems.

Ce travail est dédié au contrôle non linéaire d'un drone de type quadricoptère dont la spécificité est de pouvoir voler aussi bien que se déplacer en glissant sur le sol, à la façon d'un aéroglisseur. Dans un contexte d'exploration autonome de bâtiment, ce concept hybride permet d'économiser les batteries lorsqu'il n'est pas nécessaire de voler puisque le drone profite des surfaces planes pour se déplacer sans avoir à compenser la gravité ; il peut ainsi prolonger l'autonomie au-delà de la vingtaine de minutes typique d'un quadricoptère classique. Contrairement aux véhicules terrestres à roues, les capacités de franchissement de notre drone sont fortement augmentées car son aptitude au vol l'autorise à éviter les obstacles, à changer d'étage ou passer par une fenêtre. L'étude menée ici concerne essentiellement le déplacement surfacique de ce drone hybride, et vise à concevoir et implémenter une loi de contrôle capable d'asservir ce système sur des trajectoires planes au sol. Ce drone terrestre est similaire à un système sous actionné de type glisseur ; le problème de la stabilisation en un point est donc distingué du suivi de trajectoire en raison de la condition de Brockett que ce système ne satisfait pas ; notre plateforme ne peut donc pas être stabilisée par des retours d'états continus. En s'appuyant sur la littérature, cette thèse propose différentes approches théoriques en temps variant, fonctions transverses, platitude ou encore par "Backstepping" pour répondre à ces problèmes. Après une phase d'identification du modèle dynamique employé, la partie expérimentale, exploitant un système de Motion Capture pour récupérer les informations de position et d'orientation du système, valide ces lois de contrôle et de commande pour le suivi d'une trajectoire circulaire simple. / This thesis is dedicated to the non-linear control of a special hybrid quadrotor which is able to fly, and slide on the ground like an hovercraft. In the context of an autonomous indoor exploration this hybrid concept allows saving energy when flying is not necessary, since the drone can then slide on the ground without having to compensate for the gravity; autonomy can last beyond the 20 minutes typical of a standard quadrotor. Contrarily to wheeled mobile robots, the hybrid drone ability to move across space is strongly increased since it can fly to avoid obstacles, to move between two levels, to get in through a window. The study under consideration is essentially focused on the displacement of the drone on the ground and aims at designing and implementing a control law so that our system is able to track a 2D xy plane trajectory. This terrestrial quadrotor is similar to a slider underactuated vehicle. The point stabilisation is then separately studied from the trajectory tracking issue because of the Brockett condition, which is not satisfied in that case; our platform cannot be stabilized by means of continuous state feedbacks. This thesis proposes different theoretical developments based on the literature and deriving from time varying control laws, transverse functions, flatness or backstepping techniques to solve both point stabilisation and trajectory tracking. The experimental part of the thesis is based on the recovering of the drone position in real time and orientation via a Motion Capture system for feedback loop in the control law; the proposed dynamical model was validated as well as the control and command laws for the tracking of a circular trajectory.

Cette thèse vise à rendre possible l’utilisation des véhicules sous-marins autonomes (AUVs) dans l’inspection visuelle des structures offshore. Deux tâches sont identifiées: le suivi des structures rectilignes et la stabilisation devant les cibles planaires. Les AUVs complétement actionnés et équipés d'une centrale inertielle, un DVL, un altimètre et une caméra vidéo sont visés. La dynamique en 6 d.d.l. d'un AUV est rappelée. L'architecture de contrôle reflétant la structure en cascade de la dynamique est choisie. Une boucle interne asservie la vitesse du véhicule à la consigne et une boucle externe calcule la vitesse de référence à partir des informations visuelles. Le suivi de pipe est assuré par l'asservissement visuel 2D qui calcule la vitesse de référence à partir des bords du pipeline détectés dans l’image. La convergence globale asymptotique et localement exponentielle de la position, de l’orientation et de la vitesse sont obtenues. Le contrôleur de stabilisation utilise la matrice d’homographie. Seule la connaissance imprécise de l’orientation de la cible est nécessaire. L’information cartésienne de la profondeur de la cible est estimée à l’aide d’un observateur. La convergence quasi-globale et localement exponentielle de ce contrôleur est démontrée. Afin de tester ces méthodes un simulateur a été développé. Des images de synthèse de haute-fidélité sont générées à partir de simulateur Morse. Elles sont traitées en temps réel à l’aide de la bibliothèque OpenCV. Un modèle Simulink calcule la dynamique complète des 6 d.d.l. du véhicule simulé. Des résultats détaillés sont présentés et mettent en avant les résultats théoriques obtenus. / This thesis deals with a control system for a underwater autonomous vehicle given a two consequent tasks: following a linear object and stabilisation with respect to a planar target using an on-board camera. The proposed solution of this control problem takes advantage of a cascading nature of the system and divides it into a velocity pilot control and two visual servoing schemes. The serving controllers generate the reference velocity on the basis of visual information; line following is based on binormalized Pluecker coordinates of parallel lines corresponding to the pipe contours detected in the image, while the stabilisation relies on the planar homography matrix of observed object features, w.r.t. the image of the same object observed at the desired pose. The pilot, constructed on the full 6 d.o.f. nonlinear model of the AUV, assures that the vehicle’s linear and angular velocities converge to their respective setpoints. Both image servoing schemes are based on minimal assumptions and knowledge of the environment. Validation is provided by a high-fidelity 6 d.o.f. dynamics simulation coupled with a challenging 3D visual environment, which generates images for automatic processing and visual servoing. A custom simulator is built that consist of a Simulink model for dynamics simulation and the MORSE robot and sensor simulator, bound together by ROS message passing libraries. The OpenCV library is used for real-time image processing. Methods of visual data filtering are described. Thus generated experimental data is provided that confirms the desired properties of the control scheme presented earlier.