Research And Implementation Of Multi-agent Formation Algorithm

Multi-agent system is a system that completes complex tasks through coordination and cooperation.It is composed of multiple autonomous and mobile agents,which can greatly improve work efficiency,save costs,and reduce security risks.It is widely used in aerospace,military,transportation,disaster rescue and other fields.Multi agent cooperative control is one of the main research directions of multi-agent system.Achieving a higher level of system performance is the goal of this technology field.This article focuses on the obstacle avoidance algorithm and formation algorithm of unmanned transportation fleets in a closed environment,and tests them on the designed and built simulation and physical platforms for formation control of multiple unmanned vehicle systems.The main tasks are as follows:(1)Aiming at the problems of unreachable targets and local minima in traditional artificial potential field obstacle avoidance,an improved artificial potential field obstacle avoidance method is proposed.At the local minimum point,the weights of the gravitational and repulsive forces of the potential field are adaptively configured based on the magnitude of the gravitational and repulsive forces,thereby changing the force direction of the intelligent agent and detaching from the local point.(2)A leader predictive control algorithm based on relative position is proposed for unmanned transportation fleets in closed environments.This method introduces a model predictive controller to predict the position of the intelligent agent based on expected performance indicators and environmental constraints,improving tracking performance and solving the problem of information lag caused by communication delay.A 3D simulation platform consisting of multiple vehicles was built using Coppelia Sim,and the proposed formation control algorithm was further tested and validated,laying the foundation for system testing on the physical platform.(3)A physical experimental platform for formation control of multiple unmanned vehicle systems was designed and built,and formation control algorithms were tested.Completed the hardware and software design of the physical platform,solving key issues such as unmanned vehicle positioning,unmanned vehicle control,and communication between vehicles;Programming and implementing formation control algorithms,achieving formation control,formation change,and formation obstacle avoidance.