Motion control system analysis and design for a humanoid robot

Abstract

A humanoid robot motion control system needs to deal with information from encoders, acceleration, force, gyro and vision sensors, etc. and needs to control more than 30 degrees of freedom (DOFs) joints motion. An effective motion control system is necessary for the humanoid robot to complete walk and manipulation. This paper discusses the motion control system architecture, hardware design and control algorithm. According to the requirements of the humanoid robot, a distributed motion control system based on controller area network (CAN) bus is used in the humanoid robot system. Analysing the motion control algorithm, we used integral separation proportional integral derivative (PID) control algorithm to guarantee the humanoid robot to move and manipulate in real-time and precisely. Experiment results show that the proposed motion control system for humanoid robot is reliable and stable, and is adaptive to deal with complex tasks in daily life.