A co-evolution approach to sensor placement and control design for robot obstacle avoidance

Abstract

This paper proposes a coevolution approach to optimal sensor placement and controller design for a mobile robot to facilitate its navigation and obstacle avoidance in an unknown environment. The mobile robots considered in this paper have flexible sensor and control structure. A genetic algorithm is developed to evolve the parameters of optimal sensor placement and fuzzy logic controller simultaneously. The preliminary results indicate that the proposed coevolution approach can lead to efficient robot sensor placement and control design.