Neural-network-based formation control with collision, obstacle avoidance and connectivity maintenance for a class of second-order nonlinear multi-agent systems

摘要

In this paper, a formation control strategy with collision, obstacle avoidance and connectivity maintenance is developed for a class of second-order nonlinear multi-agent systems under external disturbances. Firstly, in order to handle the nonlinear dynamics of the multi-agent systems and the unknown disturbances in the environment, neural network (NN) techniques are employed in the proposed control strategy. Then, the distributed formation controller with collision, obstacle avoidance is designed by combining artificial potential field (APF) methods and leader & ndash;follower formation methods. Secondly, due to the collision or obstacle avoidance terms within the formation controller may result in large separation distance among agents and each agent & rsquo;s communication distance is limited by its hardware, the collision or obstacle avoidance terms increase the chance of losing connectivity between agents. In order to guarantee the connectivity of the formation, the connectivity maintenance controller is designed with taking the communication topology into account. Based on Lyapunov stability theory, it is proved that the stability of the closed-loop multi-agent systems can be guaranteed. Finally, the simulation results verify the effectiveness of the proposed approach.(c) 2021 Elsevier B.V. All rights reserved.