Strong winds are the most significant meteorological hazard that impacts the safety of high-speed trains. Strong winds can cause trains to shake in mild cases and overturn in severe cases. In order to enhance the early warning system, it is necessary to predict wind speeds and issue early warnings along the high-speed railway lines. However, there is still a gap between existing wind speed prediction models and early warning rules. Therefore, this paper proposes an Informer-Weibull method for high-speed rail wind speed prediction and early warning. Utilize Informer to forecast the extreme wind speed in the upcoming 3 min, and subsequently employ the Weibull Cumulative Distribution Function (CDF) to complete the second-level envelope within the same 3-min timeframe. The verification was completed on 137 strong wind warning events at a specific high-speed railway station. The results show that this method can effectively predict alarms and further obtain future strong wind information in advance.