Research Of Gait Planing And Motion Control For Quadruped Robot

Robot technology is a cross-discipline subject. It involves electronics, mechanics, mechanics, biology, systems engineering, computers, artificial intelligence and other disciplines of knowledge. Quadruped robot has better stability than two-legged robot, and avoids the redundancy and complexity of hexapod robot. It can walk on the rough ground and complex terrain with static gait, and also can walk quickly with dynamic gait. Quadruped robot has a wide application prospect in factory automation, construction, mining, eliminating dangerous, military, service, agriculture. It has an important practical value and social significance on the research of this aspect.The paper takes quadruped robot as research object. The research content mainly includes following aspects:1. Research four common gait of quadruped robot:crawling gait,trotting gait,pacing gait and bounding gait, and analyze the timing of various gaits. Stability margin is adopted to analyze and calculate the static stability of quadruped robot. It provides reason for the structure design and improvement of quadruped robot.2. Design the motion control hardware system structure of hierarchical structure for the quadruped robot. The control system has three layers, including control level,organization level,executive level. The entire control system consists of PC, industrial computer, motion control card, data acquisition cards and sensors. The system is an open system, it has good scalability and maintainability, reduces the development cycle of system.3. Using the intelligent algorithm of BP neural network to control the motion of quadruped robot, and simulates the algorithm with Matlab. Adopting mixed programming of VC++and Matlab to improve the programming efficiency.4. BP neural network control algorithm is verified by the walking experiment of quadruped robot, the results show that the BP neural network algorithm has good self-learning and adaptability, have some intelligence, it is feasible for the motion control of quadruped robot, it provides the hardware and theory foundation for further control law research of quadruped robot.