随机拓扑结构下多智能体系统的平均一致性

Average consensus of multi-agent system with random topology

利用马尔可夫收敛准则、图的Laplace矩阵谱特性和欧氏度量的极值,对一类具有随机拓扑结构的离散时间多智能体系统平均一致性问题进行了深入讨论。引入完好概率矩阵的概念,建立随机拓扑结构下离散时间系统的一致性算法,应用马尔可夫过程收敛相关结论及伴随算子,从欧氏度量极值的角度证明了系统可达到渐近平均一致,并得出了所需满足的条件,该条件放宽了对系统连通性的要求。最后,采用六个智能体组成的多智能体系统进行计算机仿真,对理论的正确性进行了验证。

This paper provided a theoretical framework for analysis of average consensus algorithms for multi-agent networked system with random topologies（intermittent links）.The analysis framework was based on tools from convergence of Markov process,spectral properties of graph Laplacian and the extremum of Euclidean metric.It established direct connections between properties of Markov processes and the average convergence of consensus algorithm.Simulation results are presented demonstrate the role of average connectedness of the random instantiations of the graph on the convergence of consensus algorithm.