Genetic algorithms based predictive control for mobile robot navigation in changing environments

Abstract

In this paper a predictive navigation system for mobile robots is presented. The system deals with unexpected moving obstacles which future positions over a prediction horizon are estimated with a Kalman filter approach. Genetic algorithms have been used for real time optimization problem involved in the model based predictive control problem. Experimental results obtained when applying the controller to a TRC LABMATE mobile platform are given in the paper.