Fault Line Selection Method Based on Transfer Learning Depthwise Separable Convolutional Neural Network

With the continuous development of artificial intelligence technology, the value of massive power data has been widely considered. Aiming at the problem of single-phase-to-ground fault line selection in resonant grounding system, a fault line selection method based on transfer learning depthwise separable convolutional neural network (DSCNN) is proposed. The proposed method uses two pixel-level image fusions to transform the three-phase current of each feeder into the RGB color image, which is used as the input of DSCNN. After DSCNN self-feature extraction, the fault line selection is completed. With the consideration that not all of power distribution systems can obtain a large amount of data in practical applications, the transfer learning strategy is adopted to transplant the trained line selection model. The smaller number of DSCNN parameters increases the portability of the model. The test results show that not only does the proposed method extracts obvious features, but also the line selection accuracy can reach 99.76%. It also has good adaptability under different sampling frequencies, different noise environments, and different distribution network topologies; the line selection accuracy can reach more than 97.43%.

Fault Line Selection of Non-solidly Grounding System Based on EMD and Correlation Analysis of Zero Sequence Current Break-variable

When a single-phase grounding fault occurs in non-solidly grounding system, the zero sequence transient current is nonlinear and non-stationary. Especially when the resonant system is grounded with high resistance, the transient quantity is weak, which brings challenges to data processing. Therefore, empirical mode decomposition (EMD) is proposed to decompose the transient quantity and obtain different intrinsic mode functions (IMF). The IMF with the largest discrimination is selected as the characteristic quantity for correlation analysis. At the same time, considering the existence of unbalanced current in the actual system, in order to avoid the influence of unbalanced current, a single-phase grounding fault line selection algorithm based on EMD decomposition and correlation analysis of zero sequence current break-variable is proposed. Finally, the effectiveness of the method is verified by simulation and field test waveforms.

Small Current Fault Line Selection Based on Improved Empirical Mode Decomposition and Fractal Dimension Method

In this paper, the fault steady state and transient characteristics of small current grounding system are analyzed, and the distribution of transient zero sequence current is introduced. A fault line selection based on EMD and fractal dimension method is proposed. After the parameter is determined, the problem is proposed and improved. Using the simulated annealing K-means algorithm to find the scale-free interval curve to get the line slope is the correlation dimension of the line. Finally, by comparing the size of the associated dimension, you can select the corresponding line of the faulty distribution network.