电力通信网的多层网络社团检测

Multi-layer Network Community Detection in Electric Power Communication Network

电力通信网作为电网的第二张实体网络,其可靠性以及稳定性是至关重要的,为了解决当前电力通信网部分拓扑结构鲁棒性较低问题,文章提出将基于复杂网络理论的社团检测应用到电力通信网中,可以检测出电力通信网的社团结构,通过消除社团结构,可以增强网络的鲁棒性以及稳定性。文章结合电力通信网的网络特征,提出一种多层电力通信网社团检测算法,利用业务相似度矩阵和节点相似度矩阵的拉普拉斯矩阵的特征值间隔,以及改进的K-means聚类算法实现电力通信网的社团检测,并通过与加权AA、GN等社团检测算法进行对比,实验结果表明,该文算法在模块度、准确率等指标均优于对比算法。

As the second physical network of the power grid,the reliability and stability of the power communication network are crucial.In order to solve the problem of low robustness of some topology structures in the current power communication net work,the paper proposes to apply community detection based on complex network theory to the power communication network,which can detect the community structure of the power communication network.By eliminating the community structure,the robustness and stability of the network can be enhanced.Based on the network characteristics of power communication network,the paper proposes a community detection algorithm for multi-layer power communication network,which uses the eigenvalue interval of the Laplacian matrix of the service similarity matrix and node similarity matrix and the improved K-means clustering algorithm to achieve community detection of power communication network.By comparing with weighted AA,GN and other community detection algorithms,the experimental results show that the algorithm in the paper can detect communities in the modular The accuracy and other indicators are better than the comparison algorithm.