To address the challenges posed by the complex and varied flood types in the upper and middle reaches of the Yangtze River Basin, a dynamic identification model for flood types was constructed by coupling K-Means++clustering algorithm and decision tree algorithm. The model was used to identify the types of Three Gorges inflow floods in the upper and middle reaches of the Yangtze River Basin. The results indicate that based on two indicators of the proportion of total secondary floods and the proportion of total secondary net rainfall floods,floods in the upper and middle reaches of the Yangtze River can be classified into four types, including the Min River, the Tuo River, and the Jialing River encounter type, the Min River, the Tuo River, the Jialing River, and the Xiangcun interval encounter type, the Min River, the Tuo River, the Jialing River and Xiongchun, Chunsan intervals encounter type, and the Wujiang River and Xiongchun, Chunsan intervals encounter type. The classification consistency is as high as 88.4%, and the characteristics of various types of floods are significant. The overall accuracy of dynamically identifying flood types based on rainfall information is 83.3%. The weighted average values of the accuracy, recall rate, and F1 score for different types of floods are all above 0.70, indicating that the model has good recognition performance for most flood types. Keywords: K-Means++ clustering algorithm; decision tree algorithm; flood type identification; upper and middle reaches of the Yangtze River Basin 〖FL