Research on defect detection method of transmission components based on neural network sparse coding

The reliable operation of overhead transmission lines is crucial for the safe and stable operation of the power system. In order to prevent the damage of insulators, vibration-proof hammers and other key components in transmission lines from affecting power transmission, they need to be inspected on a regular basis. In this paper, a neural network sparse coding based defect detection method for transmission components is proposed. First, the original image is decomposed at multiple scales; then, the SC method is used to learn the local features of the image and sort them according to the gray mean gradient (GMG), and the feature vectors with larger GMG are selected to initialize the convolution kernel, so that the global feature vectors are obtained through the feature learning; finally, they are dimensionality reduced by using the t-SNE method and input into a support vector machine for classification. The experimental results show that compared with the traditional defect detection methods, the algorithm in this paper has a better classification effect, and the detection accuracy and detection efficiency are significantly improved.