Stable Walking of Seven-link Biped Robot Based on CPG-ZMP Hybrid Control Method

Abstract

The Central pattern generator (CPG) method has good biomimetic effect, so it is used in the walking research of biped robot. However, the CPG control method has the disadvantage of unstable walking during the walking process of biped robot. To solve this problem, this paper proposes a hybrid control method for biped robot walking based on the CPG and Zero moment point (ZMP).Based on the physical model of the seven-link biped robot, the foot pressure sensor obtains the robot zero moment point in real time, according to the difference between it and the expected zero moment point, realizes the compensation of the CPG output joint angle, and realizes closed-loop control during walking. The parameters in the hybrid control method were optimized by GA optimization algorithm. This paper uses a self-built seven-link biped robot for simulation verification under the VREP simulation platform. The research results show that the hybrid control method can ensure a good bionic effect while achieving stable walking of the robot, and is controlled by a single CPG compared with the method, the hybrid control method can make the walking of the biped robot more stable.