Acquiring autonomous driving for a cleaning robot

Abstract

We have studied an autonomous cleaning robot by modeling it in a virtual environment, using the physical engine (PhysX). This study aims at acquiring autonomous driving and obstacle avoidance for the autonomous cleaning robot, which moves from a start place to a destination place. Autonomous cleaning robot is controlled by Artificial Neural Network (ANN), which is evolved toward learning proper actions by use of Genetic Algorithm (GA). Simulation results show that the autonomous cleaning robot acquires autonomous driving and obstacle avoidance behavior under the three dimensional virtual physical environment.