Medium-and-long-term runoff forecasting method and device considering meteorological factors and storage medium

The invention discloses a medium-and-long-term runoff forecasting method and device considering meteorological factors and a storage medium. The method comprises the following steps: acquiring influence factors (including meteorological factors and historical runoff sequences) influencing medium-and-long-term reservoir flow change based on a partial mutual information method; after the weight of each influence factor is checked, fuzzy clustering is carried out; and training a neural network by taking the clustered influence factors as input and taking medium and long term daily average runoff as output. According to the method, the linear and nonlinear relationship between the input variable and the output result can be measured, and the influence of mutual information coupling on the forecasting result can be eliminated, so that the reliability and accuracy of variable selection are ensured, and the accuracy of medium and long-term runoff forecasting is improved. 本发明公开了一种考虑气象因素的中长期径流预报方法、装置及存储介质，基于偏互信息法获得影响中