基于多传感器信息融合的水工闸门故障诊断  

Fault Diagnosis of Hydraulic Gate Based on Multi-sensor Information Fusion

作　　者：李凯旋[1,2] 张钰奇 付春健 段玥晨 LI Kai-xuan;ZHANG Yu-qi;FU Chun-jian;DUAN Yue-chen(School of Mechanical and Power Engineering,Zhengzhou University,Zhengzhou 450001,Henan Province,China;Henan Intelligent Manufacturing Research Institute,Zhengzhou 450001,Henan Province,China)

机构地区：[1]郑州大学机械与动力工程学院,河南郑州450001 [2]河南省智能制造研究院,河南郑州450001

出　　处：《中国农村水利水电》2024年第2期96-102,共7页China Rural Water and Hydropower

基　　金：工信部智能制造综合标准化与新模式应用项目(2018037);河南省水利厅水利科技攻关项目(GG202068)。

摘　　要：针对水工闸门安全检修困难、检修效率低的问题,提出了一种基于支持向量机(Support vector machine,SVM)和改进D-S证据理论的信息融合水工闸门故障诊断方法。该方法通过提取不同传感器诊断信号小波包信息熵特征构建特征子空间,然后在每个特征子空间构建诊断子网络,最后使用改进证据理论对每个诊断子网络的输入进行决策层融合,从而水工闸门的多信息融合诊断结果。闸门故障诊断实验结果显示,信息融合的闸门故障诊断方法可有效识别弧形闸门故障种类,其故障诊断准确率达到了98.33%,同时诊断可靠度高,各类故障的诊断不确定度均小于1%。实验结果验证了智能故障诊断方法用于水工闸门领域的可行性,对于改进水工闸门故障检修方式,推动水利工程智能化的发展具有重大意义。Aiming at the problems of difficult and inefficient safety maintenance of hydraulic gates,this paper proposes an information fusion fault diagnosis method for hydraulic gates based on support vector machine(SVM)and improved D-S evidence theory.The method constructs feature subspaces by extracting information entropy features of wavelet packets of different sensor diagnostic signals,and then constructs diagnostic sub networks in each feature subspace.Finally,the input of each diagnostic sub-network is fused at the decision-making level by using improved evidence theory,so as to achieve the multi-information fusion diagnosis results of hydraulic gates.The experimental results of gate fault diagnosis show that the information fusion gate fault diagnosis method can effectively identify the types of radial gate faults,with a fault diagnosis accuracy of 98.33%,and high diagnostic reliability.The diagnostic uncertainty of various faults is less than 1%.The experimental results verify the feasibility of using intelligent fault diagnosis methods in the field of hydraulic gates,which is of great significance for improving the troubleshooting methods of hydraulic gates and promoting the development of intelligent water conservancy engineering.

关 键 词：支持向量机 改进证据理论 信息融合 水工闸门 故障诊断 

分 类 号：TV663[水利工程—水利水电工程]