Research On Communication Mechanisms For Optimizing Convergence Speed Of Consensus Algorithm In ZigBee-based Multi-agent Networks

With the continuous development of science and technology and the improvement of living standards,application requirements such as UAV formation flying,multi-robot cooperative control,and large-area environmental monitoring have emerged one after another.People increasingly need solutions to such complex control problems.Under this background,multi-agent systems came into being.Multi-agent system is a complex system composed of a large number of distributed and autonomous multi-agent individuals through inter-communicating and collaborative cooperation.It has been widely used in computer networks,smart grids,smart transportation,national defense,and other fields.It is one of the fundamental issues for collaborative control to reach a consensus and realize a group objective of multi-agent networks by means of decentralized control laws,and improving convergence speed is important for consensus problems because it significantly impacts the control performance.Extensive research has been conducted on the problem of increasing the convergence speed of multi-agent systems.In the existing literature,most studies consider the improvement algorithm of convergence speed in light of the consensus protocol and network algebra connectivity.However,in fact,network communication situation is also one of the important limiting factors for the improvement of consensus convergence speed.In this paper,based on the Carrier Sense Multiple Access with Collision Avoidance mechanism under the ZigBee wireless communication protocol commonly used in multi-agent systems,the actual communication situation is modeled and analyzed,and an effective algorithm to improve the system consensus convergence speed is sought from the perspective of communication.The main work and innovations of this article are as follows:1.In order to improve the consensus convergence speed of multi-agent systems with a large number of nodes and a large amount of data transmission,this paper proposes a network access parameter adjustment algorithm based on consistency control state awareness.Firstly,based on the channel competition mechanism(Carrier Sense Multiple Access with Collision Avoid mechanism)used in ZigBee wireless communication protocol,this paper analyzes the impact of media access control layer parameters adjustment on communication characteristics;Secondly,based on the influence mechanism of parameter adjustment on communication characteristics,aiming at rationally allocating channels resources and mitigating data transmission conflicts,the basic algorithm idea of adaptive online adjustment of parameters according to the difference information of the current state value of nodes is proposed;Thirdly,according to the influence mechanism of parameter adjustment on the communication characteristics,the specific parameters used for adjustment and the amplitude and threshold of parameters adjustment are determined.Finally the specific content of the algorithm is given.The algorithm is lightweight,fully distributed and without any modification to existing wireless communication protocol.Simulation experiments show that the algorithm has faster consistency convergence and lower energy consumption compared with other algorithms.2.To further address the problem of increasing the convergence speed in the case of a multi-agent system with a large number of nodes and a larger data scale,this paper proposes an optimization algorithm for the consensus convergence speed based on the communication network partition.Firstly,network partitions can conduct to the changes not only in the degree of communication resource competition but also system algebraic connectivity.Consequently,a mathematical model between the speed of system consensus convergence and the system algebraic connectivity and the communication situation is established;Secondly,through modeling and analyzing the communication situation under the channel competition mechanism(Carrier Sense Multiple Access with Collision Avoid mechanism)based on ZigBee wireless communication protocol,the mathematical models of the successful probability of data transmission and the communication throughput in the network are obtained;Thirdly,aiming at decreasing the degree of channel competition and increasing system algebraic connectivity,the network partition algorithm which cooperates communication resource competition degree and network algebra connectivity is given and optimal number of partitions is acquired through genetic algorithm.The algorithm alleviates channel competition fundamentally through reducing the number of nodes in a partition.At the same time,it takes into account the impact of network partition on the algebraic connectivity of the system.Simulation experiments verify that the algorithm has faster consistency convergence speed and lower energy consumption compared with other algorithms.3.A combination mechanism of two algorithms mentioned before that implement network access parameters adjustment and network partition respectively to improve the consensus convergence speed is proposed.On the one hand,partition the original network offline according to the partition algorithm to find a more reasonable communication topology with lower competition of communication resources and higher algebraic connectivity of the system.On the other hand,based on the difference in the current state value of each node in the system,adjust the parameters of the media access control layer online according to the parameter adjustment algorithm to reasonably allocate existing channel resources and alleviate transmission conflicts.This combination mechanism combines the advantages of the two algorithms to increase the system consensus convergence speed preferably.