Multi-model multi-feature fusion method for predicting wind speed along high-speed railway

The invention discloses a multi-model multi-feature fusion method for predicting the wind speed along a high-speed railway, and the method comprises the following steps: 1, installing five auxiliary wind measurement stations nearby a wind measurement station; 2, carrying out the processing of original wind speed data through an interactive multi-model Kalman filtering method; 3, carrying out the wavelet processing of the filtered data, and building a prediction submodel for the low-frequency data after wavelet processing; 4, inputting target wind measurement station multistep-ahead prediction values and weather forecast target wind station prediction values obtained by a space-target wind measurement multistep-ahead prediction model, a self-target wind measurement station multistep-ahead prediction model and a weather-target wind measurement station multistep-ahead prediction model into a Bayes combined model, and obtaining the final target wind measurement station prediction value. The method can avoid the data interruption caused by a single wind measurement station hardware fault, and also can provide longer emergency processing time for the safety of the high-speed railway in a severe wind environment.