Neural sensor-based navigation of wheeled mobile robot in unknown environment

Abstract

In presented paper a new approach to a collisionfree trajectory generating for a wheeled mobile robot with Adaptive Critic Designs and Fuzzy Logic algorithm, is proposed. The presented discrete hierarchical control system consists of a trajectory generating algorithm based on a reactive navigation of the wheeled mobile robot in an unknown 2D environment with static obstacles, and a tracking control system. A strategy of reactive navigation is developed including two main behaviours: an obstacle avoiding behaviour and a goal-seeking behaviour, realised in a form of Adaptive Critic Design algorithms. These simple, individual behaviours are combined by the fuzzy combiner of behaviours that determines influence of the individual behaviours on the trajectory generation process, according to the environment conditions. The tracking control system is composed of two Dual-Heuristic Dynamic Programming algorithms, the supervisory term and the PD controller. Verification of the proposed control algorithm was realised using the mobile robot Pioneer 2-DX, equipped with one laser and eight sonar range finders, that provides object detection.