基于切换层次网络的一致性算法研究

Switching Hierarchical Network Based Consensus Algorithm Research

  • 摘要: 针对多智能体网络一致性问题,受自然界中鸽群的启发,提出一种新的有向网络拓扑即层次网络,且证明了所提出的层次网络的性能优于当前应用广泛的leader-following网络,研究了基于层次拓扑的一阶积分器网络对机动目标跟踪的一致性误差,证明一致性误差有界且与网络拉普拉斯矩阵第二小特征值及目标状态有关,并讨论了切换层次网络的一致性误差问题. 仿真结果验证了理论的有效性.

     

    Abstract: Motivated by natural pigeon flocks,a new directed topology, switching hierarchical network, was proposed to solve the consensus error of multi-agent system. It was proved that the proposed hierarchical network converges faster than leader-following network been used widely. Analyzing the consensus error generated when the network tracks a mobile target, it is showed that there is a limitary in the consensus error, the error is related to the state of target and the second smallest eigenvalue of the network Laplacian. Some problems about the consensus error based on switching hierarchical network are analyzed. The theoretic validity of the proposed switching hierarchical network is validated with simulation.

     

/

返回文章
返回