Evolutionary Artificial Neural Networks on simulated soccer robots

摘要

A soccer robot is a cooperative robot agent system on a soccer game. Robot controllers in this project must react to different environments on each robot to achieve the goal while the optimal solution of controller configuration is hard to define. The Artificial Neural Networks (ANNs) and Genetic Algorithms (GAs) give us a key to solve these problems. An ANN develops itself to process complex problem while a GA finds acceptable solutions with the ability of randomness. The project shows that these hybrid methods used in soccer league simulation yield a promising result.