Research on Fault Identification Method of Gate Meter Based on Big Data Analysis Technology
In order to improve the detection effect of gate meter metering devices in the power system and timely find the abnor-mal 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 de-tection device to judge whether the fault occurs according to the zero-sequence voltage and the zero-sequence compensation volt-age 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.
fault identificationbig data analysiscloud computingdetection devicenoise separationmemory neural network
万方数据
维普
