矩阵权重网络上混杂多智能体系统的随机一致性OA

This paper investigates the randomized consensus problem for hybrid multiagent systems on matrix-weighted networks.For such systems,this paper proposes a distributed randomized consensus protocol based on sampled information,which employs an asynchronous pairwise update mechanism to effectively reduce communica-tion and computational demand.By leveraging expected graph theory and stochastic matrix stability analysis,ne-cessary and sufficient conditions are derived for the system to achieve consensus in expectation,which are related to the sampling period and the null space of the expected Laplacian matrix.Specifically,when the feedback gains are identical,necessary and sufficient conditions are provided for achieving randomized average consensus.Furthermore,by analyzing the second-moment convergence of the error system and using Markov's inequality,an analytical up-per bound for the ϵ-consensus time is derived to characterize its convergence speed.Finally,numerical simulations validate the feasibility of the proposed protocol and the effectiveness of the theoretical results.