learning classifier systems for mobile robot control

we discuss some issues concerning the application of learning classifier system LCS to real -world applications we present the impact of using LCS applied to a mobile, autonomous robotics agent the motivation behind this approach is that classifier systems will be capable of genrating compact high performance rule sets which are genral and accurate. this paper proposes a new real time sensor based navigation method using LCS for navigation of a mobile robot in any uncertain environment. the proposed schems learns the high performance rule sets which are generalized to any workspace.

Keywords:

learning classifier systems , reinforcement learning , genetic algorithm , mobile robot