基于大数据分析技术的关口表计量故障识别方法研究  

Research on Fault Identification Method of Gate Meter Based on Big Data Analysis Technology

作　　者：马晓琴 薛晓慧 罗红郊 MA Xiaoqin;XUE Xiaohui;LUO Hongjiao(State Grid Qinghai Electric Power Company Information and Communication Company,Xining 810000,China;State Grid Qinghai Electric Power Company,Xining 810000,China)

机构地区：[1]国网青海省电力公司信息通信公司,青海西宁810000 [2]国网青海省电力公司,青海西宁810000

出　　处：《微型电脑应用》2024年第6期109-113,共5页Microcomputer Applications

基　　金：国网青海省电力公司基金(63281420008U)。

摘　　要：为了提高电力系统中关口表计量装置的检测效果,及时发现运行过程中出现的异常情况和故障,基于大数据框架设计出关口表故障识别系统,并结合云计算技术完成计算资源的虚拟化和网络资源的整合。在关口表检测装置中,主控模块使用STM32F103RBT6芯片,根据零序电压与中性点零序补偿电压判断是否出现故障。在系统的故障识别模型中结合变分模态分解和小波半软阈值分解的方法,对关口表初始故障特征信号进行处理分离噪声信号,通过长短期记忆神经网络完成故障识别。实验结果显示,该研究系统的二次压降计量误差最低为0,故障识别准确率最大为0.98,具有较好的故障检测精度,大大提升了故障识别的效率。In order to improve the detection effect of gate meter metering devices in the power system and timely find the abnormal situations and faults in the operation process,this study designs a gate meter fault identification system based on the big data framework,and combines with cloud computing technology to complete the virtualization of computing resources and the integration of network resources.The STM32F103RBT6 chip is used by the main control module in the designed gate meter detection device to judge whether the fault occurs according to the zero-sequence voltage and the zero-sequence compensation voltage of the neutral point.In the fault identification model of the system,variational mode decomposition and wavelet semi-soft threshold decomposition are combined to process the initial fault signal of the gate meter and separate the noise signal,and the fault identification is completed by the long short-term memory neural network.Experimental results show that the minimum measurement error of secondary pressure drop is 0,and the maximum fault identification accuracy is 0.98,which has a good fault detection accuracy and greatly improves the efficiency of fault identification.

关 键 词：故障识别 大数据分析 云计算 检测装置 噪声分离 记忆神经网络 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]