Estudo experimental do controle de movimento de uma plataforma Stewart do tipo 6-UPUR
Carregando...
Data
2019-07-19
Autores
Gonçalves, Amanda Hellen
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Robôs paralelos ou robôs de cadeia cinemática fechada vêm ganhando destaque no cenário industrial e acadêmico, principalmente diante da necessidade de robôs com altas acelerações e velocidades, alta relação capacidade de carga/peso e alta rigidez e precisão, motivo pelo qual são usados em simuladores de voo e robôs pick-and-place. Muitos trabalhos foram publicados tratando sobre o controle de posição de robôs paralelos, porém muitos mantiveram-se restritos ao estudo teórico, sem considerar algumas limitações na aplicação prática. Neste contexto, esta dissertação apresenta o projeto e a implementação prática de um controlador PD (Proporcional-Derivativo) independente para cada junta de um robô paralelo de seis graus de liberdade do tipo 6-UPUR (universal-prismáticauniversal- rotational) acionado por atuadores lineares eletromecânicos. Inicialmente foi realizada a calibração e o acondicionamento do sinal dos sensores de realimentação, do driver e dos atuadores eletromecânicos. Posteriormente, modelos matemáticos do comportamento dinâmico dos atuadores lineares foram identificados e, finalmente, considerando critérios de desempenho específicos para simuladores de voo, foram projetados os controladores para cada atuador. O desempenho de cada controlador foi avaliado por meio de sinais de entrada em degrau, rampa e parábola em coordenadas cartesianas e, por meio da cinemática inversa, foram calculadas as entradas desejadas para cada atuador. O desempenho do robô na frequência foi avaliado por meio de trajetórias senoidais em coordenadas cartesianas, funções descritivas e um sensor inercial de seis graus de liberdade. Os resultados obtidos mostraram as limitações do desempenho do robô, devido principalmente à saturação da amplitude da lei de controle, ruído no sensor de realimentação, zona morta e limitações de potência do atuador.
Parallel robots or closed kinematic chain robots has gained more attention from both in industry and in academic research, especially in view of the need for robot whith high velocities and accelerations, high payload-weight ratio, high stiffness and accuracy. That is why they are used in flight simulators and robots pick-and-place. Many studies have been published addressing the problem of motion control of parallel robots, but many are limited to the theoretical study, without considering some limitations in practical application. In this context, this dissertation presents the design and practical imple- mentation of independent-joint PD (Proportional-Derivative) controller for a 6-UPRU (universal-prismatic-universal-rotational) degree-of-freedom parallel manipulator driven by electromechanical linear actuators. First, the feedback sensors, drive and electrome- chanical actuators are calibrated, and their signals are processed. Later, mathematical models of the dynamical behaviour of the linear actuators are identified. Then, conside- ring specific performance criteria for flight simulators, the controller for each actuator is designed. The performance of each controller was evaluated for step, ramp and parabolic inputs in cartesian coordinates, then the cartesian trajectory is converted to desired actu- ator trajectory by using inverse kinematics. The perfomance of the robot was evaluated in frequency domain using sinousoidal inputs in cartesian coordinates, describing functions and a six degree of freedom inertial measurement unit (IMU). The results showed the limitations on the performance of the robot, mainly due to controller saturation, feedback sensor noise, dead zone and power limitations of the actuator.
Parallel robots or closed kinematic chain robots has gained more attention from both in industry and in academic research, especially in view of the need for robot whith high velocities and accelerations, high payload-weight ratio, high stiffness and accuracy. That is why they are used in flight simulators and robots pick-and-place. Many studies have been published addressing the problem of motion control of parallel robots, but many are limited to the theoretical study, without considering some limitations in practical application. In this context, this dissertation presents the design and practical imple- mentation of independent-joint PD (Proportional-Derivative) controller for a 6-UPRU (universal-prismatic-universal-rotational) degree-of-freedom parallel manipulator driven by electromechanical linear actuators. First, the feedback sensors, drive and electrome- chanical actuators are calibrated, and their signals are processed. Later, mathematical models of the dynamical behaviour of the linear actuators are identified. Then, conside- ring specific performance criteria for flight simulators, the controller for each actuator is designed. The performance of each controller was evaluated for step, ramp and parabolic inputs in cartesian coordinates, then the cartesian trajectory is converted to desired actu- ator trajectory by using inverse kinematics. The perfomance of the robot was evaluated in frequency domain using sinousoidal inputs in cartesian coordinates, describing functions and a six degree of freedom inertial measurement unit (IMU). The results showed the limitations on the performance of the robot, mainly due to controller saturation, feedback sensor noise, dead zone and power limitations of the actuator.
Descrição
Palavras-chave
Robô paralelo, Plataforma de Stewart, Controle PID, Identificação de sistemas, Controle independente das juntas, Parallel robot, Parallel robot, PID control, Independent joint control, System identification