Identification of Aging Status of Insulation Paper Based on BP Neural Network

Because of the disadvantages of the existing ways for detecting the aging state of insulation paper, this paper offers a new nondestructive testing way that makes use of 10 texture eigenvalues of the captured insulation paper images to evaluate its aging state. This paper first collects and preprocesses the images of insulation paper with different thermal aging status, and then calculates the gray scale stroke matrix of the images to obtain 10 texture eigenvalues are used as the input layer. and the corresponding thermal aging degree is used as the output layer of BP neural network. BP neural network is used to recognize the aging characteristic images of insulation paper. The results show that the error rate between the predicted value and the actual value is less than 20%, which verifies the feasibility and effectiveness of the proposed method for evaluating the aging condition of insulation paper. The method proposed in this article, combined with the telescopic endoscope, which can extend the theoretical method to practical application and bring convenience to the future transformer maintenance.