Hazardous weather such as thunderstorm, icing and turbulence is one of the main reasons for flight accidents. In order to avoid possible risk caused by temporary diversion of hazardous weather during the flight, numerical and probabilistic forecasting are used to predict enroute hazardous weather at the flight planning level. And considering the uncertainty of hazardous weather, a flight path planning method for uncertain weather is proposed, which can ensure flight operation safety to the greatest extent. Firstly, based on the probabilistic meteorological forecast data, the mapping relationship between the diagnostic elements of thunderstorms and the occurrence probability of thunderstorms is established by the mean of ingredients-based methodology from the perspective of the three occurrence conditions of thunderstorms, the water vapor condition, the instability condition and the lifting trigger condition. Then the C-F model is used to fuse the probability data of thunderstorm diagnostic elements, so that the probability of thunderstorms can be calculated and the thunderstorm area map has been set up. After that icing prediction index and turbulence prediction index are calculated based on numerical prediction data, and icing and turbulence regions were defined. Then a rasterized map that can identify the risk of hazardous weather is constructed by integrating the thunderstorms, icing and turbulence regions. On this basis, the traditional path planning algorithm is improved to maximize the flying safety, and a risk minimization Dijkstra algorithm and a risk minimization A* algorithm are proposed. Finally, by using the forecast data of a severe convective weather in Central China on April 3, 2023, a real risk map is established. Based on the map the improved risk minimization algorithms and traditional algorithms are used for path planning respectively, and the total risk and path length of each planned path are calculated for finding the minimum-risk path. The results show that the total risk of the path under risk minimization A* algorithm is minimum, the path length under traditional algorithm is minimum, and the comprehensive performance of risk minimization Dijkstra algorithm is the best. Therefore, if the flight risk is expected to be minimized, the flight path based on the risk minimization A* algorithm should be selected because the safety of that path is the highest. When the path length and risk are considered comprehensively, the flight path under the risk minimization Dijkstra algorithm can be chosen because compared with the risk minimization A* algorithm, this scheme is more economical.