Fault Diagnosis of HVDC Transmission System Using Wavelet Energy Entropy and the Wavelet Neural Network

The failure of the HVDC transmission system is the main factor affecting its reliability. There are many types of faults in the actual project. When a fault occurs, timely and effective identification of the fault type to determine the specific cause of the failure has important research value for improving the reliability of the system. Therefore, this paper focuses on the fault diagnosis method of HVDC transmission system. In this paper, a new fault diagnosis method combining wavelet energy spectrum entropy and wavelet neural network is proposed. In this method, the inverter-side converter bus voltage signal is analyzed as an electrical quantity, and the energy spectrum entropy value of the signal is used to distinguish the normal operating state from each fault state. First, the db10 wavelet is used to decompose and reconstruct the inverter-side converter bus voltage signal collected during the system operation into 10 layers and to obtain the detailed signal of wavelet reconstruction at various scales, and then calculate the wavelet energy spectrum information entropy value of each layer. Use the extracted feature energy spectrum entropy as the input feature vector of wavelet neural network, so as to realize the diagnosis of each fault type of HVDC transmission. The results show that the diagnosis method can accurately diagnose the diagnosis cause of the reduced reliability of the converter valve system.