Controle de velocidade baseado em modelo dinâmico com restrições para um robô diferencial escalador magnético

• Ano de defesa: 2021
• Autor(a) principal: Santos, Higor Barbosa
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Tecnológica Federal do Paraná; Curitiba; Brasil; Programa de Pós-Graduação em Engenharia Elétrica e Informática Industrial; UTFPR
• Palavras-chave em Português: Controle preditivo; Robôs - Sistemas de controle; Adesão; Robôs - Velocidade - Controle; Robótica; Predictive control; Robots - Control systems; Adhesion; Robots - Speed - Control; Robotics; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA::ELETRONICA INDUSTRIAL, SISTEMAS E CONTROLES ELETRONICOS::AUTOMACAO ELETRONICA DE PROCESSOS ELETRICOS E INDUSTRIAIS; Engenharia Elétrica
• Link de acesso: http://repositorio.utfpr.edu.br/jspui/handle/1/26568

Climbing robots are characterized by a secure surface coupling designed to prevent falls. The robot’s coupling capacity is ensured by an adhesion method that leads to dynamic non-linear models with time-varying parameters that affect the robot’s mobility. Besides, the friction of the wheel and the force of gravity are also relevant issues that can compromise the ability to climb if they are not well modeled. This work presents a predictive controller based on the dynamic model for speed tracking in a four-wheel magnetic climbing robot specially designed to inspect industrial storage tanks. Precise speed control is essential to guarantee the effectiveness of the inspection, which requires the compensation of dynamic disturbances present in the climbing robot’s navigation. The proposed model-based predictive controller (MPC) compensates for nonlinear effects due to inertial forces, gravity, friction, and adhesion through dynamic and kinematic modeling of the climbing robot. Dynamic modeling is based on the Lagrange-Euler approach, which allows a better understanding of how forces and torques affect the robot’s movement. Also, an analysis of the interaction force between the robot and the contact surface is carried out since this force affects the climbing robot’s movement according to the spatial orientation. As the vast majority of commercial robots use the speed of their wheels as an input signal, a transformation of torque into speed is performed, generating a new dynamic formulation for the robot’s behavior. Thus, two dynamic models for climbing robots are proposed, as the classic dynamic model uses torque as an input signal. Given the dynamic models based on torque and speed, an MPC controller is implemented for each type of input signal. Finally, simulations are performed to examine robot dynamics during the climbing movement, and MPC controllers are validated using the V-REP robot simulator and practical experiments with a real robot in a laboratory environment. The simulated and practical results of the controllers demonstrate the compensation of modeled, unmodeled and time-varying disturbances in the climbing robot navigation. From the kinematic and dynamic models developed, a good performance in the predictions was obtained, which compensated the robot’s dynamics during the climb and made the controllers present a good performance (response time and insignificant error) in the speed tracking. Therefore, the main contributions of this thesis are the kinematic and dynamic analysis and the development of predictive controllers based on the dynamic model for speed tracking of climbing robots.