Aplicações de técnica de controle para a estabilização da caminhada de uma robô bípede

Abstract

The development of mobile robotics is stimulated by the countless possibilities in the use of humanoid robots to assist humans in various tasks, such as: in housework, in hostile environments and in competitions like the RoboCup. Therefore, to deal with the most varied types of activities, it is important for robots to be able to develop locomotion strategies that are appropriate according to the environment in which they find themselves in. In this context, this work presents techniques for improving the locomotion of a humanoid robot, in order to meet the demands of the Mobile Robotics Development Team (EDROM) at UFU in competitions, applying control strategies based on the Extended Kalman filter (EKF), in order to increase the robot's locomotion stabilization. To achieve this, the robot's walking trajectory was modelled using the inverse kinematic model. The robot's trunk inclination was compensated by an inertial measurement unit (IMU), responsible for measuring the acceleration and angular velocity of the trunk and the instantaneous roll and pitch movements. With such measurements, it is possible to implement the Extended Kalman filter to reduce the noise and apply the control to the robot's walk.