Applying PSO to walking parameter tuning in RoboCup 3D Soccer Simulation

Abstract

An application of a particle swarm optimization~(PSO) to the parameter tuning problem in the bipedal walking learning for RoboCup 3D Soccer Simulation is considered in this study. Firstly, an affect of the number of particles to search performance is examined by the simulation, then we discuss the particle velocities to give any insights to improve the search performance of PSO.