Study on flood clustering and recognition methods based on fuzzy set theory

The authors give a comprehensive analysis on flood clustering and recognition methods based on fuzzy set theory. In the methods, historical flood data are clustered into several partitions by applying fuzzy clustering iteration algorithm and the optimal number of cluster is obtained by using cluster validity criterion, and a real-time flood hydrograph partition is recognized by fuzzy pattern recognition model. The above-mentioned flood clustering and recognition methods are also used to Huanren reservoir and the results show that the methods are feasible and practicable to cluster and recognize flood hydrographs.