Hydrophobic Detection Study of Composite Insulator Based on BP Neural Network Grading

Abstract

With the rapid development of industry and agriculture in our country, the atmospheric contamination degree is intensifying, which has become the main factors of influencing the electrical equipment insulation level. In order to ensure the security of system operation, the periodic detection of the hydrophobic of composite insulator in the operation is re- quired to determine its safety. According to the existing problem of low detection accuracy of the present composite insu- lator hydrophobic detection, this paper proposes a BP neural network grading based composite insulator hydrophobic de- tection model. The model is optimized and improved on the basis of the traditional BP neural network algorithm and car- ries out positive and negative transmission after input samples, then compares tonsure weight and the size of the error function, calculates according to the comparison results, adjusts the weight of nodes and results in error reduction of the traditional BP neural network algorithm. Experimental simulation results show that BP neural network grading based composite insulator hydrophobic detection model is superior to transitional BP neural network algorithm on convergence and network error, which can be well applied into the composite insulator hydrophobic detection.