Abstract
The goal of RoboCup project is that by 2050, a team of fully autonomous humanoid robot soccer players shall win the soccer game, comply with the official rule of the FIFA, against the winner of the most recent World Cup. Where RoboCup 3D Soccer Simulation is a competition in the project in order to develop robot skills to play soccer in computer simulation environment and apply the knowledge to developments for actual robots. In this competition, Robot motions are acquired by controlling angular speed of each joint. Since a fast walking motion has an advantage in soccer game, we have an interest in how to develop fast walking motions. We deal with a fast locomotion learning problem as a walking parameter tuning problem by evaluating a walking performance. The purpose of our study is to develop an efficient tuning method based on Evolutionary Computation. In the demonstration, we will explain RoboCup project, describe the walking parameter tuning problem and show a process and result of parameter tuning by using Particle Swarm Optimization.
