Frequent grounding fault location technology for multi-bamboo line pressing based on recursive bayesian algorithm

Abstract The process of locating frequent grounding faults caused by multi-bamboo line pressing in distribution networks is easily affected by the transition resistance, leading to small differences in fault modes, which makes it difficult to distinguish multiple fault scenarios. This results in poor efficiency and accuracy in fault location. To address this issue, a frequent grounding fault location technology for multi-bamboo line pressing based on the recursive Bayesian algorithm is proposed. By utilizing the transient zero-sequence equivalent circuit of frequent grounding faults caused by line pressing, the impedance factor and energy spectrum entropy attributes of the fault transient current signal are obtained. The frequent grounding faults of multi-bamboo line pressing are classified into small-angle fault mode, weak fault mode, and strong fault mode according to the value of the fault transition resistance. Corresponding Bayesian classifiers are constructed for each mode. The recursive Bayesian classifier can accurately identify the system’s fault mode. By using a waveform identification method and comparing the difference coefficients between neighboring points of adjacent waveforms, the location of the frequent grounding fault is determined. Experimental results demonstrate that this technology achieves rapid and accurate location of frequent grounding faults, with high positioning accuracy and robustness, providing a strong guarantee for the safe operation of power systems.