Grounding Fault Line Selection in the Distribution Network Using a Novel Underdetermined Blind Source Separation Algorithm

When resonant grounding occurs, the fault components of the line are complex, and components with different properties cancel out or compensate each other, weakening the fault characteristics and easily making it difficult to select the grounding fault line. Considering the reality of line measurement channels, this paper utilizes an underdetermined blind source separation algorithm to separate the steady-state component, transient component, and noise of faults from the measurement signal. Using the high-order moments of each component of the zero-sequence current of the line as the criterion, accurate line selection of single-phase grounding faults can be achieved. Compared with Prony, HHT, and FFT methods, the method proposed in this paper has certain advantages. The simulation and on-site experimental results show that the accuracy of the method in line selection is all correct, and it has certain anti-interference ability.