Fuzzy Clustering Method Based on Set Pair Analysis and Optimal Fuzzy Equivalent Matrix for Flood Classification

摘要

In order to reflect different importance of each index in flood classification, fuzzy analytic hierarchy process was used to determine the weights of flood classification indexes with reason, set pair analysis was used to construct fuzzy similar matrix R, and then a fuzzy clustering method based on Set Pair Analysis and Optimal Fuzzy Equivalent Matrix, named SPA-OEEM for short, was established. The applied results show classification results are reasonable and high precision, the optimal fuzzy equivalent matrix Q can improve matrix transitive degree of R and reduce distortion problem of transitive closure method, and clustering results of history flood samples can be obtained directly based on Q. SPA-OFEM is a simple and effective improved fuzzy clustering method; it can be widely applied in management of flood disaster.