Research on Thermal-Acoustic Fusion Diagnosis of GIS Equipment Defects Based on Improved D-S Evidence Theory

Due to the fully enclosed characteristics of GIS equipment and its internal defects are difficult to detect, overheating and partial discharge is a common defect within the GIS. If hidden defects cannot be detected and eliminated in time, insulation equipment will accelerate defect degradation, resulting in insulation breakdown or surface flashover in GIS equipment. Aiming at the two typical defects of GIS internal electrical contact and point partial discharge, the external transmission mechanism of internal defects was studied separately, and the detection method of defect features was explored in this paper. In the process, using improved Kalman filtering algorithm to denoise the external thermal and acoustic signal features of equipment defects, and using support vector machine algorithm to predict the internal fault point features of the equipment from external information. An information fusion algorithm based on improved D-S evidence theory was investigated to synthesize thermal and acoustic signal features to discriminate the defect status level of equipment. Finally, a GIS single branch bus defect simulation experiment platform was built to verify the effectiveness of the thermal-acoustic information fusion algorithm proposed. After fusing the defect information features of six detections, the algorithm can accurately discern the equipment defect type and status level.