Method for Identifying Abnormal Hot Spots on the Surface of Composite Insulators

【Objective】To avoid the operational risk of high-voltage transmission lines caused by thermal faults of composite insulators, a method for identifying abnormal hot spots on the surface of composite insulators is studied.【Methods】Three types of abnormal surface heating forms of composite insulators, current heating type, voltage heating type, and comprehensive heating type, are analyzed. Based on the analysis results, the fiber Bragg grating centralized temperature measurement technology is used to deploy the fiber Bragg grating sensor near the composite insulator. The fiber optic dispersion centralized temperature measurement unit is used to calculate the corresponding wavelength offset and temperature change value. After being analyzed by the digital signal processing unit, the thermal image of the composite insulator is generated. Using the maximum inter class variance method to segment the thermal image of composite insulators, the average gray level of the thermal image of composite insulators is calculated, and the optimal threshold to partition the pixels in the thermal image of composite insulators is obtained. Based on the thermal image segmentation results of composite insulators, pixel statistical methods are used to calculate the horizontal row pixel points of the composite insulator segmentation image and divide the structural area of the composite insulator. According to the application specification of thermal image diagnosis for live equipment and the principle of fiber grating centralized temperature measurement, the thermal image of composite insulator collected by the fiber grating sensor is displayed in pseudo color to show the surface temperature distribution of composite insulator. The abnormal hot spots on the surface of composite insulators are identified through the temperature relationship among various structural regions of the target image.【Results】The experimental results show that this method can accurately measure the temperature of any point on the surface of composite insulators, and the maximum Root-mean-square deviation is only 1.39%. It is shown that the proposed method can effectively divide the structural regions of composite insulator segmentation images, with clear boundaries between each structure and no overlap phenomenon.【Conclusion】The recognition performance of abnormal hot spots on the surface of composite insulators in different areas of high-voltage transmission lines is relatively ideal, and the division effect of composite insulator structure areas is relatively good.