Fault Diagnosis Strategy of Substation Electrical Equipment Based on Image Segmentation

Due to the different forms of equipment fault in substation, the required treatment methods are also different, and there is no applicable standard treatment method. Aiming at the problems in infrared detection, this paper analyzes the structure, fault types and infrared thermal image characteristics of several kinds of key equipment in substation. Then, the infrared image standard processing flow which is suitable for these kinds of equipment is proposed, and the corresponding optimization is made according to the specific structure characteristics of each kind of equipment to realize the segmentation and extraction of target area. A convolution neural network model is constructed to realize the recognition of infrared image. Finally, an infrared detection method of substation key equipment based on image recognition and neural network is proposed. The experimental results show that the proposed infrared image processing process and the established fault recognition network and image classification network have high practicability, and they are expected to further promote the application of infrared detection work.