An Comprehensive Detection Method for Fault Location and Fire Risk Judgment of Substation Power Equipment

Abstract

In view of the lack of fault identification and fire risk prediction of power equipment detection in substation, an comprehensive detection method for fault location and fire risk judgment is proposed in this paper. The maximum information entropy theory is used to deal with the fuzziness of infrared image to improve the imaging contrast. The Gaussian mixture modeling and temporary background updating are used to locate suspicious faults. The temperature and area of remote and small target are identified by subpixel irradiance analysis. The temperature rise and fire risk characteristics are constructed and the equipment defect and fire risk levels are defined, which is to realize the advance warning of equipment abnormality and fire risk. This method is applied in the equipment inspection of a 500kV substation to verify the effectiveness and practicability of the method.