The main focus of this study is on the active and semi-active control of civil engineering structures subjected to seismic excitations. Among different candidate control strategies, the sliding mode control approach emerges as a convenient alternative, because of its superb robustness under parametric and input uncertainties. The analytical developments and numerical results presented in this dissertation are directed to investigate the feasibility of application of the sliding mode control approach to civil structures. In the first part of this study, a unified treatment of active and semi-active sliding mode controllers for civil structures is presented. A systematic procedure, based on a special state transformation, is also presented to obtain the regular form of the state equations which facilitates the design of the control system. The conditions under which this can be achieved in the general case of control redundancy are also defined. The importance of the regular form resides in the fact that it allows to separate the design process in two basic steps: (a) selection of a target sliding surface and (b) determination of the corresponding control actions. Several controllers are proposed and extensive numerical results are presented to investigate the performance of both active and semi-active schemes, examining in particular the feasibility of application to real size civil structures. These numerical studies show that the selection of the sliding surface constitutes a crucial step in the implementation of an efficient control design. To improve this design process, a generalized sliding surface definition is used which is based on the incorporation of two auxiliary dynamical systems. Numerical simulations show that this definition renders a controller design which is more flexible, facilitating its tuning to meet different performance specifications. This study also considers the situation in which not all the state information is available for control purposes. In practical situations, only a subset of the physical variables, such as displacements and velocities, can be directly measured. A general approach is formulated to eliminate the explicit effect of the unmeasured states on the design of the sliding surface and the associated controller. This approach, based on a modified regular form transformation, permits the utilization of arbitrary combinations of measured and unmeasured states. The resulting sliding surface design problem is discussed within the framework of the classical optimal output feedback theory, and an efficient algorithm is proposed to solve the corresponding matrix nonlinear equations. A continuous active controller is proposed based only on bounding values of the unmeasured states and the input ground motion. Both active and semi-active schemes are evaluated by numerical simulations, which show the applicability and performance of the proposed approach. / Ph. D.

Ein Haupteinsatzgebiet der Großraummanipulatoren stellen Betonverteilermasten dar. Aufgrund der langen schmalen Armkonstruktionen fällt bei diesen Maschinen der Trend zum Leichtbau bezüglich der Dynamik besonders ins Gewicht. Um die Vorteile leichter Konstruktionen wie geringere Achslasten, geringerer Kraftstoffverbrauch und kleinere Antriebe nutzen zu können, werden Regelungen benötigt, die die Struktur stabilisieren und ein Schwingen der Mastspitze verhindern. Zur Systemanalyse und Regelungsauslegung wurde ein Mehrkörpermodell aus starren und elastischen Körpern sowie den notwendigen Hydraulikzylindern erstellt und durch Messungen validiert. Am Modell konnte gezeigt werden, dass die Regelung im letzten Gelenk die Schwingung an der Mastspitze maßgeblich beeinflusst und zur Schwingungstilgung eingesetzt werden kann. Hierfür wird die Bewegung des Verteilermastes durch eine Ausgleichsbewegung im letzten Gelenk kompensiert, sodass die Mastspitze keine starken Schwingungen ausführt. Die Schwingungen werden über Beschleunigungsmessung detektiert und nach entsprechender Filterung kann die Bewegung bestimmt werden. Mittels Sliding Mode Control erfolgt die Berechnung der schwingungsmindernden Zylinderkraft und garantiert somit Robustheit gegenüber Modellierungsungenauigkeiten und äußeren Störungen. Die Kraftregelung des Hydraulikzylinders wird anschließend über eine Integrator-Backstepping Regelung realisiert. Die resultierende Schwingungsminimierung beträgt in unterschiedlichsten Maststellungen bis zu 95%. / A special case of multi-link manipulators are truck mounted concrete pumps. Due to the lightweight design of the long and slender boom, it is vulnerable to vibrations. The advantages are smaller masses and therefore less actuation power which results in smaller actuators with less fuel consumption. In order to retain the advantages of lightweight design, special controllers are needed to stabilize the overall system and result in a vibration free motion of the boom tip. A multibody system with flexible bodies has been built in order to analyse the system's behaviour and to test and design appropriate control strategies. It could be demonstrated, that controlling only the last joint of the boom decisively effects the motion of the boom tip and is therefore suitable to suppress vibrations. The idea is to compensate the boom's motion by adjusting the last joint angle in a way, so that the boom tip stays at its initial position. In order to implement these findings and obtain a robust control three steps are necessary: the boom's motion must be measured and a vibration reducing force defined which has to be applied by the hydraulic actuator. The vibrations are detected by acceleration measurement and after appropriate filtering a joint angle trajectory can be determined. The cylinder force is found using Sliding Mode Control which guarantees robustness against modeling inaccuracies and external disturbances. A mathematical description of the last segment is necessary for the design of this nonlinear control strategy. The force control of the hydraulic cylinder is then implemented via backstepping control. The resulting vibration is minimized by this control by up to 95% at different boom positions.

Ein Haupteinsatzgebiet der Großraummanipulatoren stellen Betonverteilermasten dar. Aufgrund der langen schmalen Armkonstruktionen fällt bei diesen Maschinen der Trend zum Leichtbau bezüglich der Dynamik besonders ins Gewicht. Um die Vorteile leichter Konstruktionen wie geringere Achslasten, geringerer Kraftstoffverbrauch und kleinere Antriebe nutzen zu können, werden Regelungen benötigt, die die Struktur stabilisieren und ein Schwingen der Mastspitze verhindern. Zur Systemanalyse und Regelungsauslegung wurde ein Mehrkörpermodell aus starren und elastischen Körpern sowie den notwendigen Hydraulikzylindern erstellt und durch Messungen validiert. Am Modell konnte gezeigt werden, dass die Regelung im letzten Gelenk die Schwingung an der Mastspitze maßgeblich beeinflusst und zur Schwingungstilgung eingesetzt werden kann. Hierfür wird die Bewegung des Verteilermastes durch eine Ausgleichsbewegung im letzten Gelenk kompensiert, sodass die Mastspitze keine starken Schwingungen ausführt. Die Schwingungen werden über Beschleunigungsmessung detektiert und nach entsprechender Filterung kann die Bewegung bestimmt werden. Mittels Sliding Mode Control erfolgt die Berechnung der schwingungsmindernden Zylinderkraft und garantiert somit Robustheit gegenüber Modellierungsungenauigkeiten und äußeren Störungen. Die Kraftregelung des Hydraulikzylinders wird anschließend über eine Integrator-Backstepping Regelung realisiert. Die resultierende Schwingungsminimierung beträgt in unterschiedlichsten Maststellungen bis zu 95%. / A special case of multi-link manipulators are truck mounted concrete pumps. Due to the lightweight design of the long and slender boom, it is vulnerable to vibrations. The advantages are smaller masses and therefore less actuation power which results in smaller actuators with less fuel consumption. In order to retain the advantages of lightweight design, special controllers are needed to stabilize the overall system and result in a vibration free motion of the boom tip. A multibody system with flexible bodies has been built in order to analyse the system's behaviour and to test and design appropriate control strategies. It could be demonstrated, that controlling only the last joint of the boom decisively effects the motion of the boom tip and is therefore suitable to suppress vibrations. The idea is to compensate the boom's motion by adjusting the last joint angle in a way, so that the boom tip stays at its initial position. In order to implement these findings and obtain a robust control three steps are necessary: the boom's motion must be measured and a vibration reducing force defined which has to be applied by the hydraulic actuator. The vibrations are detected by acceleration measurement and after appropriate filtering a joint angle trajectory can be determined. The cylinder force is found using Sliding Mode Control which guarantees robustness against modeling inaccuracies and external disturbances. A mathematical description of the last segment is necessary for the design of this nonlinear control strategy. The force control of the hydraulic cylinder is then implemented via backstepping control. The resulting vibration is minimized by this control by up to 95% at different boom positions.

This thesis relates to the development of the Ambidextrous Robot Hand engineered in Brunel University. Assigned to a robotic hand, the ambidextrous feature means that two different behaviours are accessible from a single robot hand, because of its fingers architecture which permits them to bend in both ways. On one hand, the robotic device can therefore behave as a right hand whereas, on another hand, it can behave as a left hand. The main contribution of this project is its ambidextrous feature, totally unique in robotics area. Moreover, the Ambidextrous Robot Hand is actuated by pneumatic artificial muscles (PAMs), which are not commonly used to drive robot hands. The type of the actuators consequently adds more originality to the project. The primary challenge is to reach an ambidextrous behaviour using PAMs designed to actuate non-ambidextrous robot hands. Thus, a feasibility study is carried out for this purpose. Investigating a number of mechanical possibilities, an ambidextrous design is reached with features almost identical for its right and left sides. A testbench is thereafter designed to investigate this possibility even further to design ambidextrous fingers using 3D printing and an asymmetrical tendons routing engineered to reduce the number of actuators. The Ambidextrous Robot Hand is connected to a remote control interface accessible from its website, which provides video streaming as feedback, to be eventually used as an online rehabilitation device. The secondary main challenge is to implement control algorithms on a robot hand with a range twice larger than others, with an asymmetrical tendons routing and actuated by nonlinear actuators. A number of control algorithms are therefore investigated to interact with the angular displacement of the fingers and the grasping abilities of the hand. Several solutions are found out, notably the implementations of a phasing plane switch control and a sliding-mode control, both specific to the architecture of the Ambidextrous Robot Hand. The implementation of these two algorithms on a robotic hand actuated by PAMs is almost as innovative as the ambidextrous design of the mechanical structure itself.

Orientador: Marcelo Carvalho Minhoto Teixeira / Banca: Rodrigo Cardim / Banca: Alfredo Del Sole Lordelo / Resumo: Este trabalho trata de sistemas incertos com controle com modos deslizantes. O foco principal das contribuições é dado na investigação deste método de controle realizado através de rede de comunicação sujeitas a atrasos. Para minimizar os efeitos degenerativos dos atrasos, preditores de estado no seu formato mais simples são utilizados na lei de controle. São feitas análises da robustez da estabilidade e também da influência que as incertezas e atrasos exercem sobre o comportamento destes sistemas em malha fechada. Para corroborar os resultados apresentados nas análises, são realizadas simulações em um modelo matemático de um sistema de ordem dois, em um modelo linear do sistema de suspensão ativa e em um modelo não linear do sistema pêndulo invertido. Neste último, o atraso é tratado como uma falha e apresenta-se uma estratégia para detectar e adaptar os controladores a este tipo de falha. Também são investigados, através de simulações, aplicações do controle com modos deslizantes em dois processos integrantes do tratamento primário de petróleo / Abstract: This work deals with control of uncertain systems with sliding modes. The main focus of the contributions is given in the investigation of this method of control accomplished via communication network subject to time delay. To minimize the degenerative effects of delay, state predictors in its simpler format is used in the control law. Analyses of the stability robustness are made and also analyses about the influence that the uncertainties and delay exert on the behavior of closed-loop systems. To corroborate the results presented in the analyses, simulations are performed on a mathematical model of a system of order two, in a linear model of the active suspension system and a nonlinear model of the inverted pendulum system. In this last, the delay is treated as a failure and a strategy to detect and adapt the controllers to this type of failure is presented. Through simulations, sliding mode control in two processes of the primary oil treatment is also investigated / Mestre

Este trabalho apresenta a aplicação da teoria de controle robusto não linear por modos deslizantes a sistemas de posicionamento dinâmico para embarcações flutuantes, com validação experimental. O objetivo do sistema de controle projetado é manter a embarcação próxima a uma posição pré-ajustada (set-point) ou a uma trajetória preestabelecida (pathfollowing), por meio das forças geradas nos propulsores, mesmo estando o sistema na presença de distúrbios externos, ou seja, vento, ondas e correnteza. A princípio, realizaram-se simulações numéricas com o sistema projetado a fim de verificar o seu desempenho. O simulador utilizado foi implementado em ambiente Matlab/Simulink, considerando a dinâmica da embarcação e dos agentes ambientais. As simulações consistiram de manobras realizadas em condições nominais e na ausência de esforços ambientais, com embarcação cheia (plena) e vazia (lastro). Para validação do algoritmo implementado realizaram-se ensaios de manobra em condição de calmaria e na presença de vento, com a embarcação em plena carga e vazia. Os ensaios foram administrados no laboratório do Departamento de Engenharia Naval e Oceânica da USP (DENO). O algoritmo de controle por modos deslizantes demonstrou-se robusto a variações de condições ambientais (vento), mantendo o desempenho e estabilidade. Verificou-se que o ajuste dos parâmetros do controlador pode ser feito de forma intuitiva, utilizando-se fórmulas matemáticas. Além disso, a estrutura não linear do controlador e suas propriedades de robustez asseguram o desempenho e estabilidade para uma grande gama de condições ambientais e manobras realizadas com a embarcação. / This paper presents the application of the robust and nonlinear sliding mode control theory to the dynamic positioning systems for floating vessel, with experimental validation. The objective of the control system designed is to keep the vessel next a specific position (set-point) or follow a pre-defined trajectory (pathfollowing) through the action of propellers, in the presence of wind, waves and current external disturbances. In principle numerical simulations were carried out with the system designed to verify its performance. The simulator used was implemented in a Matlab / Simulink, considering the dynamics of the vessel and environmental agents. The simulations consisted of maneuvers carried out in nominal condition and in the absence of environmental efforts, with the vessel full and empty (ballasted). In order to validate the algorithm, small scale experiments were done, considering maneuvers in both calm and windy conditions, with the vessel at full or ballasted load. The tests were conducted at the laboratory of the Naval and Ocean Engineering Department (DENO) of the University of São Paulo. The sliding mode control was robust to variations in environmental conditions (wind), keeping the performance and stability. It was verified that the adjustment of controller parameters can be easily done, using mathematical equations. Moreover, the nonlinear structure of the controller and its robustness properties ensure the performance and stability for a large range of environmental conditions and maneuvers carried out with the vessel.

"This thesis presents the development of a magnetic resonance imaging (MRI) compatible pneumatic actuation test system regulated by piezoelectric valve for image guided robotic intervention. After comparing pneumatic, hydraulic and piezoelectric MRI-compatible actuation technologies, I present a piezoelectric valve regulated pneumatic actuation system consisted of PC, custom servo board driver, piezoelectric valves, sensors and pneumatic cylinder. This system was proposed to investigate the control schemes of a modular actuator, which provides fully MRI-compatible actuation; the initial goal is to control our MRI-compatible prostate biopsy robot, but the controller and system architecture are suited to a wide range of image guided surgical application. I present the mathematical modeling of the pressure regulating valve with time delay and the pneumatic cylinder. Three different sliding mode control (SMC) schemes are proposed to compare the system performance. Simulation results are presented to validate the control algorithm. Practical tests with parameters determined from simulation show that the system performance attained the goal. A novel MRI- compatible locking device for the pneumatic actuator was developed to provide safe lock function as the pneumatic actuator fully stopped."

Cette thèse s’inscrit dans le cadre du projet "Petits drones dans le vent" porté par le centre ONERA de Lille. Ce projet vise à utiliser le drone comme "capteur du vent" pour gérer un quadcopter UAV dans des conditions aérologiques perturbées en utilisant une prédiction du champ de vent. Dans ce contexte, le but de la thèse est de faire du quadcopter un capteur de vent pour fournir des informations locales afin de mettre à jour le système de navigation. Grâce à l’estimation du vent à bord en temps réel, le quadcopter peut calculer une planification de trajectoire évitant les zones dangereuses et le contrôle de trajectoire correspondant basé sur une cartographie existante et doté des informations relatives au concernant le comportement aérodynamique de l’écoulement d’air à proximité des obstacles. Ainsi, les résultats de cette thèse, dont les objectifs principaux portent sur l’estimation du vent instantanée et le contrôle de position, seront fusionnés avec une autre étude traitant de la planification de trajectoire. Un problème important est que les capteurs de pression, tels que l’aéroclinomètre et le tube de Pitot, ne sont pas facilement utilisables à bord des véhicules à voilure tournante car l’entrée des rotors interfère avec le flux atmosphérique et les capteurs LIDAR légers généralement ne sont pas disponibles. Une autre approche pour estimer le vent consiste à mettre en œuvre un logiciel d’estimation (ou un capteur intelligent). Dans cette thèse, trois estimateurs de ce type sont développés en utilisant l’approche du mode glissant, basée sur un modèle de drone adéquat et des mesures disponibles sur le quadcopter et sur des systèmes de position de suivi inertiel. Nous nous intéressons ensuite au contrôle de la trajectoire également par mode glissant en considérant le modèle non linéaire du quadcopter. Nous étudions par ailleurs de façon encore assez préliminaire une solution alternative fondée sur la commande H, en considérant le modèle linéarisé pour différents points d’équilibre en fonction de la vitesse du vent. Les algorithmes de contrôle et d’estimation sont strictement basés sur le modèle détaillé du quadcopter, qui met en évidence l’influence du vent / This thesis is part of the project "Small drones in the wind" carried by the ONERA center of Lille. This project aims to use the drone as a "wind sensor" to manage a UAV quadrotor in disturbed wind conditions using wind field prediction. In this context, the goal of the thesis is to make the quadrotor a wind sensor to provide local information to update the navigation system. With real-time on-board wind estimation, the quadrotor can compute a trajectory planning avoiding dangerous areas and the corresponding trajectory control, based on anexisting cartography and information on the aerodynamic behavior of airflow close to obstacles. Thus, the results of this thesis, whose main objectives are to estimate instant wind and position control, will be merged with another study dealing with trajectory planning. An important problem is that pressure sensors, such as the aeroclinometer and the Pitot tube, are not usable in rotary-wing vehicles because rotors air inflow interferes with the atmospheric flow and lightweight LIDAR sensors generally are not available. Another approach to estimate the wind is to implement an estimation software (or an intelligent sensor). In this thesis, three estimators are developed using the sliding mode approach, based on an adequate drone model, available measurements on the quadrotor and inertial tracking position systems. We are then interested in the control of the trajectory also by sliding mode considering the nonlinear model of the quadrotor. In addition, we are still studying quite an early alternative solution based on the H control, considering the linearized model for different equilibrium points as a function of the wind speed. The control and estimation algorithms are strictly based on the detailed model of the quadrotor, which highlights the influence of the wind

Este trabalho apresenta a aplicação da teoria de controle robusto não linear por modos deslizantes a sistemas de posicionamento dinâmico para embarcações flutuantes, com validação experimental. O objetivo do sistema de controle projetado é manter a embarcação próxima a uma posição pré-ajustada (set-point) ou a uma trajetória preestabelecida (pathfollowing), por meio das forças geradas nos propulsores, mesmo estando o sistema na presença de distúrbios externos, ou seja, vento, ondas e correnteza. A princípio, realizaram-se simulações numéricas com o sistema projetado a fim de verificar o seu desempenho. O simulador utilizado foi implementado em ambiente Matlab/Simulink, considerando a dinâmica da embarcação e dos agentes ambientais. As simulações consistiram de manobras realizadas em condições nominais e na ausência de esforços ambientais, com embarcação cheia (plena) e vazia (lastro). Para validação do algoritmo implementado realizaram-se ensaios de manobra em condição de calmaria e na presença de vento, com a embarcação em plena carga e vazia. Os ensaios foram administrados no laboratório do Departamento de Engenharia Naval e Oceânica da USP (DENO). O algoritmo de controle por modos deslizantes demonstrou-se robusto a variações de condições ambientais (vento), mantendo o desempenho e estabilidade. Verificou-se que o ajuste dos parâmetros do controlador pode ser feito de forma intuitiva, utilizando-se fórmulas matemáticas. Além disso, a estrutura não linear do controlador e suas propriedades de robustez asseguram o desempenho e estabilidade para uma grande gama de condições ambientais e manobras realizadas com a embarcação. / This paper presents the application of the robust and nonlinear sliding mode control theory to the dynamic positioning systems for floating vessel, with experimental validation. The objective of the control system designed is to keep the vessel next a specific position (set-point) or follow a pre-defined trajectory (pathfollowing) through the action of propellers, in the presence of wind, waves and current external disturbances. In principle numerical simulations were carried out with the system designed to verify its performance. The simulator used was implemented in a Matlab / Simulink, considering the dynamics of the vessel and environmental agents. The simulations consisted of maneuvers carried out in nominal condition and in the absence of environmental efforts, with the vessel full and empty (ballasted). In order to validate the algorithm, small scale experiments were done, considering maneuvers in both calm and windy conditions, with the vessel at full or ballasted load. The tests were conducted at the laboratory of the Naval and Ocean Engineering Department (DENO) of the University of São Paulo. The sliding mode control was robust to variations in environmental conditions (wind), keeping the performance and stability. It was verified that the adjustment of controller parameters can be easily done, using mathematical equations. Moreover, the nonlinear structure of the controller and its robustness properties ensure the performance and stability for a large range of environmental conditions and maneuvers carried out with the vessel.

Orientador: Cláudio Kitano / Abstract: Interferometers are extremely sensitive measurement devices, which use the principle of interference between two or more sources of light to generate a pattern of constructive and destructive interference. This pattern contains information about the physical phenomenon under study, and their light intensity can be used to calculate the optical path difference traveled by the two beams. The optical path difference and light intensity relationship is given by a cosine type function. Large disturbances can change the interferometer operation point, reaching nonlinear regions of the interferometric curve and even inducing ambiguities due to the periodicity of the input/output relationship. The present work concerns with the modeling, development and application of a control strategy based on sliding mode control, in a two-beam quadrature interferometer. It was used the high gain approach, which consists in to fully compensate the phase shifts induced on the sensor arm with the control system, in such a way that the voltage control signal becomes proportional to the phase disturbances. Therefore, the demodulation process does not require phase unwrapping algorithms. This implemented system showed capability to improve dynamic range and bandwidth when compared with other control systems in literature that were based on different high gain approach topologies. Also a new method of interferometric phase demodulation is proposed allying this control strategy to a virtual emulated inte... (Complete abstract click electronic access below) / Resumo: Interferômetros são dispositivos de medição extremamente sensíveis, os quais utilizam o princípio de interferência entre duas ou mais fontes de luz para gerar um padrão de interferência construtiva e destrutiva. Este padrão contém informação sobre os fenômenos físicos sob estudo, e sua intensidade luminosa pode ser usada para calcular a diferença de caminho óptico acumulada pelos dois feixes de luz. A diferença de caminho óptico e a intensidade de luz são relacionadas por uma função cossenoidal. Grandes distúrbios podem alterar o ponto de operação do interferômetro, alcançando regiões não lineares da curva característica do interferômetro e até mesmo induzindo ambiguidades, devido à periodicidade da relação entrada/saída. Este trabalho preocupou-se com o modelamento, desenvolvimento e aplicação de uma estratégia de controle baseada em controle com modos deslizantes, em um interferômetro de dois feixes em quadratura. Foi utilizada a abordagem de alto ganho, a qual consiste em utilizar o sistema de controle para compensar completamente os deslocamentos de fase induzidos no braço sensor, de tal forma que o sinal de controle se relaciona com os deslocamentos de fase por uma equação de reta. Portanto o processo de demodulação não necessita de algoritmos de desdobramento de fase. O sistema implementado mostrou capacidade de melhorar a faixa dinâmica e largura de banda quando comparado com outros sistemas de controle na literatura, também baseados na abordagem de alto ganho. Destaca... (Resumo completo, clicar acesso eletrônico abaixo) / Doutor