A PSO algorithm for biped gait planning using spline approximation

Abstract

this paper discusses an optimal trajectory planning method for a compass gait biped robot. The biped robot is composed of a stance leg and a swing leg. Each step of walking locomotion is divided into two phases, i.e., the swing phase and toe collision phase. It is assumed that the toe collision is perfect plastic and occurs instantly. The motion of the swing and stance legs in swing phase is solved by the optimal trajectory planning based on the particle swarm optimization (PSO) and the spline approximation method. The performance index function is designed as the integral of weighted sum of input torque's square, each joint trajectory is approximated by a spline function. The optimal trajectory solutions are obtained by PSO method. The hip joint can be approximately passive by increasing the corresponding weighted factor, so that a smaller motor can be used at the hip joint to reduce the joint mass. Computer simulations are performed for the optimal trajectory planning method for a compass gait biped robot. Simulation results show that the proposed method is more energy-effective than the virtual gravity method.