RBF神经网络的关口计量装置故障诊断方法研究  

Research on Fault Diagnosis Method for Gateway Metering Device Based on RBF Neural Network

作　　者：刘月骁 陆翔宇 李蕊 杨广华 李娜 LIU Yuexiao;LU Xiangyu;LI Rui;YANG Guanghua;LI Na(State Grid Beijing Electric Power Research Institute,Beijing 102600,China)

机构地区：[1]国网北京电科院,北京102600

出　　处：《电工技术》2022年第20期133-136,139,共5页Electric Engineering

基　　金：低压电流互感器计量性能评估技术研究项目(编号5700-202111201A-0-0-00)。

摘　　要：为解决目前关口计量装置面临的快速有效针对关口计量装置故障进行预警及计量装置故障类型判别困难两个主要问题,提出一种基于RBF神经网络的关口计量装置故障分析方法。该方法利用RBF神经网络算法对采集的电力计量二次回路数据进行目标对应训练,并利用MATLAB仿真平台进行实例测试,结果显示基于RBF神经网络诊断模型可以达到很好的预警和识别效果。该模型在关口计量装置故障诊断领域具备前沿性和应用价值,可以进一步研究和推广。In order to solve the two main problems faced by the current gateway metering device:fast and effective early warning for the failure of the gateway metering device and difficulty in distinguishing the type of metering device failure,a fault analysis method of gateway metering device based on RBF neural network is proposed.The RBF neural network algorithm is used to carry out target corresponding training on the collected secondary circuit data of power metering,and the MATLAB simulation platform is used for example testing.The results show that the diagnostic model based on RBF neural network can achieve good early warning and recognition effects.The model has cutting-edge and application value in the field of fault diagnosis of gateway metering devices,and can be further studied and promoted.

关 键 词：关口计量装置 RBF神经网络 故障诊断模型 电力计量 

分 类 号：TM933[电气工程—电力电子与电力传动]