In the past few years, transmission line galloping accidents have occurred frequently, seriously threatening the safe operation of the power grid. In order to achieve the purpose of early warning and early prevention, and reduce the impact of transmission line galloping on the safe operation of power grid, an early warning method of transmission line icing and galloping based on data mining was proposed. Firstly, the transmission line icing growth model was established, the BP neural network was used to analyze the weight of various meteorological factors of the conductor icing, and on the basis of the meteorological numerical prediction results and topographic information, the Bayes-Adaboost method was used to model the galloping probability of the transmission line, which realizes the early warning of the transmission line galloping probability. Finally, the method was verified according to the historical meteorological data of Henan power grid, and the effectiveness of the model was proved.