Reactive navigation of a mobile robot using elliptic trajectories and effective online obstacle detection

Abstract

This paper deals with the problem of mobile robot navigation in cluttered environment. Adaptive elliptic trajectories are exploited for reactive obstacle avoidance using only position information and uncertain range data. The obstacle avoidance strategy used is based on the elliptic limit-cycle principle where each obstacle is surrounded by an ellipse. The ellipse parameters are computed online using a sequence of uncertain range data. An online heuristic method combined with the extended Kalman filter (EKF) is used to compute the ellipse parameters. It is demonstrated that this process ensures that all range data are surrounded by a computed ellipse. Moreover, this paper proposes a single control law to the multicontroller architecture where a reactive obstacle avoidance algorithm is embedded. The proposed control law is based on the Kanayama control law; it is designed to improve the performance of the controllers. The stability of this control architecture is proved according to the Lyapunov synthesis. Simulations and experiments in different environments have been performed to demonstrate the efficiency and reliability of the proposed online navigation in cluttered environment.