Enabling holistic insulation monitoring in secondary AC protection circuits: a diagnostic algorithm

This paper addresses the problem of insulation status monitoring in secondary alternating current (AC) circuits of power system relay protection. A comprehensive diagnostic algorithm based on multi-dimensional feature fusion is proposed. This algorithm integrates distributed zero-sequence current monitoring, waveform similarity analysis, and third harmonic analysis techniques to construct a multi-dimensional feature fusion diagnostic model. By real-time acquisition of zero-sequence current in each branch, combined with the similarity changes of current waveforms under normal and abnormal operating conditions and the characteristics of third harmonic content, early identification and accurate location of typical faults such as insulation degradation, single-point grounding, and multi-point grounding are achieved. Experimental results show that the algorithm achieves a 98.6% detection rate for insulation faults, a 94.3% accuracy rate for multi-point grounding location, an average diagnostic latency of only 43.63 ms, and a false alarm rate of 1.5% under normal operating conditions. This significantly improves the timeliness and accuracy of insulation fault diagnosis, providing effective technical support for the safe and reliable operation of secondary circuits in substation relay protection.