Experience repository based Particle Swarm Optimization for evolutionary robotics

摘要

In this paper, experience repository based Particle Swarm Optimization (ERPSO) is proposed for effectively applying Particle Swarm Optimization (PSO) to evolutionary robotics application. The ERPSO uses a concept experience repository to store previous position and fitness of particles to accelerate convergence speed of PSO. We applied the ERPSO to find parameter of gait of a quadruped robot that produces fast gait and ERPSO showed best performance among original PSO and PSO variants. ERPSO has fast convergence property which reduces the evaluation of fitness of parameters in a real environment.