Research On Key Technologies Of Formation And Obstacles Avoidance In Multi-agent System

MAS(Multi-agent System)is a set composed of multiple agents.The goal is to break a complex System into multiple small agents,agents can communicate and coordinate with each other,which has become an important branch of distributed artificial intelligence.Formation obstacle avoidance technology of multi-agent system is a technology that enables the multi-agent system to form and maintain a certain formation according to the distributed task requirements or in the process of tracking the target and avoid obstacles in the process of movement under specific constraints.Formation obstacle avoidance technology can be used in military missions to carry out area pusher demining in complex environments,so as to improve the efficiency and safety of demining missions.At the same time,this technology is also widely used in cargo handling and robot football matches.Multi-agent formation obstacle avoidance technology can effectively improve the performance of distributed system and reduce system cost,which has practical significance and research value.This paper took multi-agent formation obstacle avoidance technology as the research background.First on multi-agent formation technology,based on the analysis of common formation methods,the leader-follower method was studied in detail,which is simple and easy to implement and has good adaptability to complex environments,but has two defects: the robustness of the formation is highly dependent on the robustness of the leader and the movement of the follower cannot affect the movement of the leader.Therefore,a formation control method combining virtual structure method and leader-follower method was proposed in this paper.This method introduces the concept of rigid body into the formation structure model of leader-follower method,and adopts polling method to solve the problems caused by leader-follower failure.Two kinds of formations,triangle and linear,were designed,and the simulation results showed that the formation had good stability and robustness.Secondly for multi-agent system obstacle avoidance technology,based on the analysis of common obstacle avoidance algorithms,?-FVF(?-Fuzzy and Virtual Force)algorithm was proposed.This algorithm combines Mamdani fuzzy control theory with influence factor and virtual force field method to effectively solve the problem of unreachable target and local optimal solution in obstacle avoidance planning.The simulation results showed that the algorithm is efficient and the obstacle avoidance path is smooth and reliable.Finally,on the basis of the above research results,in view of the multi-agent formation system in complex obstacle problem of dynamic formation transformation was analyzed.The zero transformation,the scale transformation and the complete transformation were defined,and the pattern of formation transformation was specified by the discriminant function of formation transformation.The formation transformation performance and environmental adaptation evaluation function were designed to optimize the formation selection.Simulation results showed that this strategy is effective.