基于IIARA-CSA-ELM算法的WSN节点故障诊断模型  

Fault Diagnosis Model of WSN Nodes Based on IIARA-CSA-ELM Algorithm

作　　者：吴佩霖 何涛 王红卫 齐放 谭俊 WU Peilin;HE Tao;WANG Hongwei;QI Fang;TAN Jun(Information Communication Company,State Grid Hubei Electric Power Company,Wuhan 430077,China)

机构地区：[1]国网湖北省电力有限公司信息通信公司,湖北武汉430077

出　　处：《电力信息与通信技术》2022年第6期89-97,共9页Electric Power Information and Communication Technology

摘　　要：为提高无线传感器网络(wireless sensor networks,WSN)节点故障诊断效率,文章提出一种基于改进的归纳属性约简算法(improved inductive attribute reduction algorithm,IIARA)和使用乌鸦搜索算法(crow search algorithm,CSA)优化后的极限学习机(extreme learning machine,ELM)构建的诊断模型。首先使用IIARA算法对WSN故障诊断决策表进行约简;然后针对ELM稳定性和精确性偏低的问题,引入CSA算法对ELM的输入权值和隐含层阈值进行优化;最后构建出IIARA-CSA-ELM模型实现对WSN节点故障的准确识别与分类。仿真结果证明,该模型在5种不同可靠性的数据集中,均能够达到较高的诊断效率,有效提升了WSN节点故障诊断水平。In order to improve the efficiency of wireless sensor networks(WSN)node fault diagnosis,this paper proposes a diagnosis model based on improved inductive attribute reduction algorithm(IIARA)and extreme learning machine(ELM)optimized by crow search algorithm(CSA).Firstly,IIARA algorithm is used to reduce the WSN fault diagnosis decision table.Then,aiming at the problem of low stability and accuracy of ELM,CSA algorithm is introduced to optimize the input weight and hidden layer threshold of ELM.Finally,the IIARA-CSA-ELM model is constructed to realize the accurate identification and classification of WSN node faults.Simulation experiments are carried out with five data sets with different reliability.The results show that the model can achieve more prominent diagnosis performance and effectively improve the fault diagnosis level of WSN nodes.

关 键 词：WSN 节点故障 粗糙集 归纳属性约简算法 乌鸦搜索算法 ELM 

分 类 号：TN929.5[电子电信—通信与信息系统] TP212.9[电子电信—信息与通信工程]