Protective pad identification method for lightweight operation and maintenance of smart substation

At present, the operation and maintenance of protective pad in smart substation are still mainly manual in order to improve the work efficiency. This paper proposes a new online identification method. Firstly, the neural network algorithm is used to predict the state of protective pad, then the forward reasoning framework is used to optimize the protective pad prediction model, and finally the model is transplanted to the mobile terminal, and the cell phone software is used to complete the protective pad state calibration. The method was experimentally verified, and the experimental results show that the accuracy of the method reaches more than 99%, the average recognition speed reaches 40 pictures, and can effectively solve the influence of environmental factors such as reflection and occlusion, which can significantly improve the operation and maintenance efficiency of smart substation.