Research on Fault Line Selection Based on Information Fusion For Small Current Grounding System

Abstract

Aiming at the problem of low accuracy of the single line selection method for small-current ground faults, this paper proposes a line selection method based on transient and steady-state information fusion. First, the zero-sequence signal is processed by empirical mode decomposition and Fast Fourier Transform to extract the three fault feature quantities of the IMF energy, the fifth harmonic component, and the active power component; then the concept of the fault measure is introduced, and the fault measure function is used to calculate Three fault measures of feature quantities are used, and fault measures are used as feature input vectors for information fusion; finally, least squares support vector machine (LSSVM) classifiers are used as information fusion line selection algorithms to classify the model accuracy and generalization. On the problem of the parameter selection that has a greater impact on the capability, the particle swarm optimization algorithm (PSO) is used to optimize the parameters, and the PSO-LSSVM fault line selection model is established using the optimized parameters. By modeling the small current grounding system in PSCAD and using MATLAB to process data, the results show that the model improves the accuracy of fault line selection.