Prediction of transmission line icing is a key technology to ensure the normal operation of transmission lines in Beijing during the icing season. Aiming at Beijing area, Pearson correlation coefficient and grey system correlation degree analysis method were used to study the correlation between ice cover thickness and micrometeorological and microtopography based on historical data, and the results showed that humidity, slope direction, wind direction and elevation had a higher degree of influence on ice cover thickness; The ice cover prediction model based on the extremely randomized trees model and the grey system prediction model was built by combining various environmental characteristics. The root mean square error (RMSE) of the prediction results of different models was compared, and the grey system prediction model based on the combination of humidity and wind direction was the best. The results show that, compared with the similar prediction methods, the influence of micro-topography on icing thickness prediction is considered, and the environmental factors with high correlation of icing thickness of transmission lines in Beijing are humidity, aspect, wind direction and elevation. Compared with the icing prediction models constructed by various environmental factors, the root mean square error of the grey system prediction model combining humidity and wind direction is significantly better than other combinations, which can effectively realize the icing prediction of transmission lines in Beijing.