Icing Image based power system voltage stability evaluation of transmission lines

Abstract

Many transmission lines are in high-altitude and icing areas, facing the risk of line breaking and tower falling. However, the traditional manual inspection method has slow identification speed and low identification accuracy, resulting in a lot of labor cost. In this paper, an image recognition method of transmission line voltage stability risk assessment based on icing image is proposed to realize fast and accurate identification of transmission line voltage stability risk. Firstly, the transmission line icing image data are collected to identify the risk of transmission line break. Then, a convolutional neural network model based on Mobilenet-V3 framework is built, and the transmission line icing image data is used as the input and transmission line break risk as the output of the model. The model is trained to generate a rapid assessment model of transmission line voltage stability risk. Then, another deep learning model is built which can achieve the fast calculation of voltage stability margin of the system based on the different transmission line break. And the product of line breaking risk and voltage stability margin obtained from the Mobilenet-V3 model and deep learning model is taken as the final voltage stability evaluation result. Finally, the model is tested on a 500kV transmission line in a province. The test results show that this method can realize the rapid and accurate voltage risk assessment of transmission line.