Fault Diagnosis and Condition Division Criterion of DC Gas Insulating Equipment Based on SF 6 Partial Discharge Decomposition Characteristics

SF 6 decomposition characteristics directly reflect the generation, development, and evolution of internal faults in gas-insulated equipment (GIE). However, establishing the DC-GIE internal insulation condition comprehensive division criterion and fault diagnosis method based on SF 6 decomposition characteristics remains unsolved. Therefore, partial discharge (PD) and SF 6 decomposition characteristics of typical insulation defects from initial discharge to the breakdown process are discussed in this paper. The average discharge magnitude in a second $Q_{\text {sec}}$ is selected as an electrical feature to characterize the PD condition, and the PD state is divided into three levels, namely, mild PD, medium PD, and severe PD. SF 6 decomposition experiments are conducted under different PD conditions of each defect, and the SF 6 decomposition characteristics under different failure conditions are obtained. On this basis, we construct five feature sets of 36 characteristic quantities for GIE PD fault diagnosis and analyze the validity. A minimum-redundancy-maximum-relevance-based feature selection algorithm is used to optimize the characteristic quantities, and an optimal feature set is selected, i.e., {C(SO 2 )/C(T), C(SOF 2 )/C(CT), C(SO 2 F 2 )/C(ST)}. The characteristic quantities of SF 6 decomposition component, which characterizes PD condition, are reconstructed, and the component characteristic criterion of PD condition division is proposed. Thus, a decision tree evaluation model for the PD condition of the typical insulation defect inside DC-GIE is constructed. This paper lays a foundation for the insulation condition evaluation and fault diagnosis of DC-GIE.