Research For Cooperative Control Of Multi-agent Systems Based On Distributed Network In Complex Environments

In recent years,research on distributed collaborative control of multi-agent systems has attracted more and more researchers' attention.It's not only because it can reveal the characters of many complex systems in nature,more importantly,there is a wide engineering application background of it.However,in the process of cooperative control of multi-agent systems,along with the local information interaction,the communication delay of the distributed network arises naturally,and affects the performance of the system,and then destroys the stability of the system.In the past few years,the research focus of academic circles has turned to the research about the consistency of multi-agent systems with the constraints of network delay.However,the existing research results are not based on the specific network model,and that means there is not important practical significance.Based on classic application scenarios,we carry out collaborative design of ne twork protocols and multi-agent system control mechanism,we have made improvements in the following aspects,including:(1)Considering the distributed network resource scheduling protocol in the context of poor spectrum resources,an information interac tion delay model based on the unmanned aerial vehicle group coordinated control service characteristics is established,and a robust delay mechanism design is carried out on this basis.Compared with the existing results,we comprehensively consider the distributed network protocol,consistent coordination and regulation mechanism,and application scenarios to design the system globally,which effectively improves the limitations of the original theory.(2)Considering the fast-changing battlefield environment's requirements for convergence speed performance,based on the network's forwarding capability and distributed network resource scheduling mechanism,a second-order consistency regulation mechanism based on neighbor state information in a two-hop range is proposed.The stability and convergence performance of the hop consistency control mechanism under the constraints of distributed networks.(3)we comprehensively consider the constraints of the specific application scenario,the working mechanism of the collaborative system,the resource scheduling protocol of the distributed network,and the robust performance requirements of the delay.A two-hop predictive consensus formation protocol for a multi-agent collaborative system based on a distributed network is proposed.We theoretically analyze the stability of the algorithm and complete the simulation of the algorithm's convergence performance.(4)We set up a distributed network in laboratory conditions to complete the experimental verification.The results show that the system can achieve convergence to the expected formation under the condition that the parameter design meets the theoretical range,and its convergence performance is better than the classical algorithm.Therefore,this is a consistent and coordinated regulatory mechanism for drone groups that adapts to complex environmental constraints.Based on the above research results,we have completed the design of the coordinated and coordinated control mechanism of the multi-agent system that adapts to the constraints of complex environments.