降雨空间分布的模糊熵聚类分析

为了提高降雨量插值精度,充分挖掘降雨变量信息,利用模糊熵聚类分析算法,对流域内雨量站进行模糊熵聚类研究,通过基于可能性分布和距离判定的聚类有效性函数确定模糊熵系数和聚类数,从而得到模糊聚类结果,改进原有的插值方法。以淮河流域蚌埠站以上区域99个雨量站雨量数据,分别在一般情况下和模糊熵聚类情况下做交叉验证,结果显示,模糊熵聚类分析在反距离平方插值法中对降雨精度有所提升。

Fuzzy entropy clustering analysis of spatial distribution of precipitation

In order to improve the accuracy of precipitation interpolation and fully explore the information regarding precipitation variables, fuzzy entropy clustering(FEC)was carried out at rain gauge stations in a basin. A clustering validity function, based on possibility distribution and distance determination, was used to determine the fuzzy entropy coefficient and the number of clusters, so as to obtain the fuzzy clustering results and improve the original interpolation method. Based on data from 99 rain gauge stations located above the Bengbu Station in the Huaihe River Basin, cross validation was conducted under non-clustering and FEC conditions. The results demonstrate that FEC improves the precipitation accuracy in the inverse distance squared interpolation method.