【摘要】 考虑降雨时空分布、降雨平均强度和土壤初始含水量对洪水过程的影响,将这些影响因子分为数值特征影响因子和属性特征影响因子,运用模糊C均值算法对历史洪水样本模糊聚类,然后根据不同的类建立不同的洪水预报模型。实例验证表明,该方法可以有效提高洪水预报的精度。更多还原

【Abstract】 The flood characteristic influence factors were divided into the value characteristic influence factors and the attribute characteristic influence factors in this paper.In the identical kind of flood,the influence on the flood of these two kind of characteristics influence factors is the same.Thus,the characteristic weighted fuzzy clustering algorithm(FCM) was used to classify the history flood sample,and then the parameters of each kind of the Xin’an River model were optimized with each kind of sample separately.The application shows that the method presented in the article may improve the precision of the flood forecast effectively.更多还原