摘要
为消除指标相关性对聚类造成的影响,避免主观地选取汛期时间域,提出综合方法进行汛期分期。使用变异系数法和灰色关联法对所选指标进行筛选,利用集对分析法划分汛期时间域,最后利用Fisher最优分割法进行汛期分期。对陈家湾水库进行汛期分期,得出结果:汛期为6月1日—10月5日;前汛期为6月1日—6月25日,主汛期为6月26日—9月5日,后汛期为9月6日—10月5日。所得结果与前人分析结果较为一致,表明该方法符合实际,有一定的适用性。
In order to eliminate the influence of index correlation on clustering,and avoid subjective selection of flood season,a comprehensive method for flood season staging was proposed.The selection index was selected by using the coefficient of variation method and grey correlation method,and the time domain of flood season was divided by set pair analysis.Finally,the flood season was divided by Fisher optimal segmentation method.The results show that the flood season is from June 1st to October 5th,the former flood season is from June 1st to June25 th,the main flood season is from June 26th to September 5th,and the latter flood season is from September 6th to October 5th.The results are consistent with the previous analysis results,which show that the method is practical and has certain applicability.
引文
[1]王丽学,孙晓东,李学森,等.汛期分期定量分析方法在汤河水库中的应用研究[J].水电能源科学,2011,29(3):44-47.
[2]郭金城,郭倩,武学毅.基于模糊集合分析法与圆形分布法的水库汛期分期研究[J].水电能源科学,2013,31(3):50-53.
[3]侯玉,吴伯贤,郑国权.分形理论用于洪水分期的初步探讨[J].水科学进展,1999,10(2):140-143.
[4]罗姗姗,宋星原,张艳军,等.汛期分期分析方法在佛子岭水库群的应用研究[J].水电能源科学,2012,30(2):44-48.
[5]朱惠倩.聚类分析的一种改进方法[J].湖南文理学院学报(自然科学版),2005,17(3):7-9.
[6]肖聪,顾圣平,崔巍,等.Fisher最优分割法在李仙江流域汛期分期中的应用[J].水电能源科学,2014,32(3):70-74.
[7]杨逾,郑志明.基于范例推理的软基处理评价方案研究[J].长江科学院院报,2015,32(12):93-97.
[8]王文圣,李跃清,金菊良,等.水文水资源集对分析[M].北京:科学出版社,2010:27-32.
[9]王贺佳,武鹏林.基于Fisher最优分割法的汛期分期[J].人民黄河,2015,37(8):30-34.
[10]刘克琳,王银堂,胡四一,等.Fisher最优分割法在汛期分期中的应用[J].水利水电科技进展,2007,27(3):14-16.
[11]李俊,武鹏林.改进的Fisher最优分割法在汛期分期中的应用[J].中国农村水利水电,2016(11):23-26.
[12]WU P L,LI J,WANG J Y.Floods Season Division Based on Improved Set Pair Analysis[J].International Conference on Applied Mechanics and Materials,2016(10):29-30.