Support Vector Machine Based Dynamic Obstacle Avoidance Strategy for First-Order Consistent Formation Control

This paper addresses the challenge of dynamic obstacle avoidance in wheeled mobile robots by developing a classifier grounded in multi-agent consensus theory, leveraging support vector machines to enhance obstacle avoidance efficiency. Firstly, the current main methods of dynamic obstacle avoidance were introduced. Based on the existing consensus algorithm, support vector machine was selected as the dynamic obstacle avoidance method. Subsequently, specific attackers are designated to target both the leader and the follower, followed by simulations of the formation system to derive significant conclusions. That is, support vector machine can have good performance in multi-agent obstacle avoidance, which proves the feasibility and effectiveness of support vector machine in multi-agent dynamic obstacle avoidance. The simulation results show that in the limit case, that is, when the dynamic obstacle avoidance target continuously attacks the formation, the consensus formation system can still maintain the formation and recover to the original formation.