Formation Control and Obstacle Avoidance for Multi-agent Systems in Unknown Environment

This paper addresses the formation control and the obstacle avoidance problems for multi-agent systems in unknown environment with obstacles. On the basis of the above problems, two control algorithms are proposed. Firstly, a behavior-based algorithm is proposed for single-agent to plan a collision-free path in concave obstacle environment. In this algorithm, we define three behaviors of the agent, namely, obstacle avoidance behavior, ignoring the repulsive force behavior, and moving along obstacle behavior. The terminal conditions of each behavior are designed to guarantee that the agent switches the behavior reasonably according to the surrounding information and its own state. Secondly, an algorithm for formation obstacle avoidance is presented. The algorithm makes the formation avoid colliding with obstacles and reach the goal position by planning the virtual leader's path. Simulation results show the effectiveness of the proposed control algorithms.