Control System Design And Implementation Of Biped Robots

Biped robot is a combination of computer science, electronics, communications, control science, sensors, and other technologies. It represents the highest achievement in mechanical and electrical integration, and becomes the hot research area in technology. Control system is the heart of the robots, which determines the performance of the robots. Biped robot haves more joints and complicate structure, with more requirement in walking speed, self-adapting. In this case, control system design of biped robot faces more challenge. This paper analyzes the control structure of biped robots and presents a modularized robot control system based on CANopen and RTLinux. The biped robot which adopted this control system structure performs very well in the experiment.First, this paper reviews and summarizes the biped robot research history and development status. We analyze and compare the biped robot designed by different research institution, and give a detailed introduction on the control system design of these robots.Secondly, we analyze the control system design proposal, according to the mechanical structure of the robot. Represent the hardware structure based on CAN and give a introduction of the robot’s communication, sensing and motor control system. The CANopen-based application layer communication protocol is also introduced, which implement the coordinated movement of all joints.Thirdly, this paper presents a RTLinux-based software architecture for biped robot, considering the real-time performance of the robot. Give a detailed introduction of RTLinux device driver design and real-time application design. We designed a modularized software structure according to the biped robot requirement and analyzed its real-time performance.Finally, we did many experiments on biped robots such as crouch, walking forward and walking backward. In the experiments, the robot walked smoothly and harmoniously. We analyze the experiment results in the paper and give some advice on improving the robot control system.