Topology Control And Optimization Of Multi-agent System Network Based On Constraint Set

With the continuous development of modern distributed theory and distributed technology,the amount of information that needs to be processed is also increasing.The efficiency and cost of centralized methods can no longer meet people's needs.Therefore,distributed multi-agent systems(MASs)came into being.The first consideration of a distributed multi-agent system is the issue of consensus,that is,to allow all members of the system to reach a consensus on one or more state values.The communication and cooperation of MASs rely on a communication network composed of various nodes.In an agent network with large-scale nodes,communication will be very complex and redundant,which is not conducive to the system's consensus.Therefore,network topology optimization of MAS is very important.In a multi-agent system with limited sensing range,the constraint set can always keep each agent in constant communication with all its neighbors,thus ensuring global connectivity.However,when the topology has a large coverage area,selecting all neighbors to calculate the convergence effect of the constraint set is not ideal.Maintaining connectivity and enabling the system to quickly converge on large-scale high-density multi-agent systems has always been a huge challenge.This paper mainly studies multi-agent systems with a discrete time lower limit and a sensing range,optimizes the classic consensus protocol,and proposes three consensus protocols based on optimized networks,which fully reflect the characteristics of the multi-agent system's timing and effectively enhance for consensus.In the multi-agent system with limited sensing range,the constraint set can be able to keep the communication between each agent and all neighbors at every time from beginning to end,thus ensuring the global connectivity.However,selecting all neighbors to calculate the constraint set is unsatisfactory with convergence when a topology with large coverage.Maintaining connectivity and enabling systems to converge rapidly on a largescale and high-density multi-agent system has been a research challenge.In the network connectivity protection algorithm based on network boundary,a concept of network topology network edge is proposed.The system uses the distributed algorithm BCP(Boundary Connectivity Preservation)to enable each agent to find the network boundary locally and maintain its connectivity,and finally achieve the effect of maintaining the global boundary.At the same time,DMA(Dimension Maximum Average Algorithm)is designed,which can reduce the impact of the uneven distribution of agents on the system.More importantly,DMA ensures that the agent will be relatively orderly with its neighbors in each iteration,and will not exceed the local network boundary,ensuring global connectivity.On the other hand,the BCP algorithm can reduce the communication edge by about 80%.The greater the topological density,the more obvious the effect.In order to design a more general connectivity preservation algorithm.In the multiagent network connectivity protection algorithm based on density subgraph,a distributed SDB(Sector Division Based)consensus protocol is proposed to accelerate convergence,and a density-based maintain multi-agent system connectivity DSG(d-Subgragh)algorithm.First,the SDB algorithm allows each agent to divide its neighbors into different sectors,choose different strategies to calculate its control input and limit its upper limit according to the distribution.Then,based on the local density information,DSG constructs multiple connected components that can maintain connectivity during one iteration.Each agent only maintains the connectivity with certain neighbors through the constraint set according to the local connected components,so as to protect the global connectivity.This method can be applied to multi-agent systems of different dimensions,and is easier to understand and implement,and can have a wider range of applications.A motif-based multi-agent consensus protocol was first proposed.First,multi-agent consensus protocol usually proceeds over the low-order structure by only considering the direct edges between agents,but ignores the higher-order structure of the whole topology network.Second,the existing work assumes all the edges in a topology network have the same weight without exploring the potential diversity of the connections.In this way,multi-agent systems fail to enforce consensus,resulting in fragmentation into multiple clusters.To address the above issues,this paper proposes a Motif-aware Weighted Multiagent System(MWMS)method for consensus control.We focus more on triangle motif in the network,but it can be extended to other kinds of motifs as well.First,a novel weighted network is used which is the combination of the edge-based lower-order structure and the motif-based higher-order structure,i.e.,hybrid-order structure.Subsequently,by simultaneously considering the quantity and the quality of the connections in the network,a novel consensus framework for MAS is designed to update agents.Finally,we have proved the effectiveness and efficiency of our method through theoretical proofs or a large number of comparative simulation experiments,and the connectivity protection algorithms mentioned are independent of consistency and they have strong reuse.It can be applied to the different problem of multi-agent systems with limited sensing ranges.