Integrated Model for Recognition of Single-Phase-to-Ground Fault Caused by Insulation Deterioration in Medium-Voltage Distribution Networks

Abstract

Insulation deterioration is one of the key causes of the single-phase-to-ground fault (SPGF) in medium-voltage distribution networks, and accurate recognition of the SPGFs caused by insulation deterioration can improve the maintenance efficiency and prevent the development of faults. An integrated model for recognition of SPGFs caused by insulation deterioration is proposed in this paper. Multiple waveform features in the time, frequency, and time-frequency domains are extracted, and their feasibility is analyzed through a multivariate analysis of variance. Then, recognition models based on the fuzzy inference system, extreme learning machine, and support vector machine are designed, and an integrated recognition model based on Dempster-Shafer evidence theory is proposed. The recognition result of onsite data indicate that the integrated model outperforms any single model, confirming the advantages and feasibility of the proposed model.