CAJ | 학술논문

由于缺少公认的标准，洪灾评估带有很大的主观性，同时，当样本集中存在超大样本或过小样本时，会导致模糊综合评估结果出现较大偏移。提出了一种基于模糊聚类迭代模型的洪灾评估方法，无需提供标准即能对样本空间进行智能辨识并作出相应处理，并以四川省1976-2006年洪灾样本为例，给出了四川省洪灾的等级评估结果。与投影寻踪方法的评估结果进行比较表明，该方法可以适用于大部分情况下的洪灾评估工作。
유우결소공인적표준，홍재평고대유흔대적주관성，동시，당양본집중존재초대양본혹과소양본시，회도치모호종합평고결과출현교대편이。제출료일충기우모호취류질대모형적홍재평고방법，무수제공표준즉능대양본공간진행지능변식병작출상응처리，병이사천성1976-2006년홍재양본위례，급출료사천성홍재적등급평고결과。여투영심종방법적평고결과진행비교표명，해방법가이괄용우대부분정황하적홍재평고공작。
Due to lack of generally accepted standards, flood disaster assessment is often with a great deal of subjectivity. Mo-reover, when there are maximal and minimal samples in the sample set, the fuzzy comprehensive evaluation result may deviate significantly. This paper proposes a flood disaster assessment method based on fuzzy clustering iterative model without providing standard. This method can intelligently identify the sample space and make a corresponding process on it. Taking 1976-2006 flood disaster samples in Sichuan Province as an example, the flood grade assessment results are given and compared with the projection pursuit method. The assessment results show that this method can be applied in most cases of flood assessment work.