Mechanical defects diagnosis for gas insulated switchgear using acoustic imaging approach

Mechanical faults are major fault types in gas insulated switchgear (GIS). The vibration and acoustic signature provide valuable health indicator for GIS. This paper proposes a novel mechanical defect detection and identification approach for GIS using acoustic imaging. The vibration modals of a three-phase-in-one-tank GIS bus tube under different natural frequencies are investigated. An acoustic pressure sensor matrix is designed to obtain the acoustic intensity nephogram of the GIS. The gray-level co-occurrence matrix (GLCM) is used to extract the characteristics of the nephogram. Five eigenvalues, including angular second moment, contrast, entropy, correlation, and inverse different moment, are analyzed for three types of mechanical defects (i.e., loose anchor bolts, loose disc insulator bolts, and poor switchgear contact). A mechanical defect diagnostic approach for GIS is developed using the variation percentage of the averaged values of contrast as the key indicator. The effectiveness of the approach is validated using experimental data.