Dynamic obstacle avoidance based on time-variation of a potential field for robots formation

This paper presents a new dynamic obstacle avoidance strategy based on time-variation of a potential field, with multiple control objectives. The strategy uses the null space of a Jacobian matrix to achieve the different control objectives in a non-conflicting way while using either a flexible or a rigid formation to avoid static and mobile obstacles. By modifying the priorities of control objectives without changing the controller structure. Overall control system stability is analyzed and proven through Lyapunov theory. Experimental results for a three-robot formation show the performance of the proposed controllers.