Abstract: :In order to improve the pollution flashover prevention level of power system, this paper takes the assessment method of pollution state on insulator surface of overhead transmission lines in Beijing as the research object, and proposes a dynamic intelligent assessment method of regional pollution degree based on meteorological conditions and salt density data distribution. First, this paper analyzes the main external factors affecting insulator surface pollution, and uses grey correlation analysis to select six factors with the highest correlation degree. Secondly, a prediction model of insulator pollution degree based on long short-term memory neural network is established in this paper. The prediction results of this model are in good agreement with the measured data, which can well reflect the pollution state of insulators in power grid. Finally, by combining meteorological conditions with geographic information, this paper draws the forecast map of pollution area distribution in Beijing with ArcGIS, and realizes the dynamic expression of pollution degree evaluation. Moreover, a miniature natural fouling platform is built on campus to verify the proposed method. The results show that the research work in this paper can realize the periodic dynamic change monitoring of pollution degree data of regional transmission grid, predict the occurrence location of heavy pollution areas in advance, and help to improve the predictability, accuracy and intelligence level of anti-pollution flasticity work of power grid.

Key words: transmission line external insulation, meteorological conditions, grey correlation analysis, long short-term memory(LSTM), pollution degree prediction model, ArcGIS