| K均值聚类分析方法在洪水预报中的应用 |
| |
| 引用本文： | 钱堃,包为民,李偲松,司伟.K均值聚类分析方法在洪水预报中的应用[J].水电能源科学,2012,30(5):41-44. |
| |
| 作者姓名： | 钱堃 包为民 李偲松 司伟 |
| |
| 作者单位： | 河海大学水文水资源与水利工程科学国家重点实验室,江苏南京210098;河海大学水文水资源学院,江苏南京210098 |
| |
| 基金项目： | 国家自然科学基金资助项目(40901015,41001011);高等学校学科创新引智计划基金资助项目(B08048);教育部科技基础性专项基金资助项目(2007FY140900);教育部“长江学者和创新团队发展计划”基金资助项目(IRT0717) |
| |
| 摘 要： | 针对在洪水预报中历史信息未得到充分利用的问题，采用K均值聚类分析方法对历史洪水进行聚类，并分类进行新安江模型的参数率定，通过计算洪水指标到各聚类中心的距离来判别即将发生洪水的归属类别，根据判别结果采用对应类的参数进行预报。结果表明，该方法扩大了信息的利用量，提高了洪水预报的精度。
|
| 关 键 词： | 洪水预报 K均值聚类 模型参数 分类 |
Application of K-mean Cluster Analysis in Flood Forecasting |
| |
| QIAN Kun,BAO Weimin,LI Caisong and SI Wei.Application of K-mean Cluster Analysis in Flood Forecasting[J].International Journal Hydroelectric Energy,2012,30(5):41-44. |
| |
| Authors: | QIAN Kun BAO Weimin LI Caisong and SI Wei |
| |
| Affiliation: | State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China; College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China;State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China; College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China;State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China; College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China;State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China; College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China |
| |
| Abstract: | Aiming at the problem of not fully considering the history information in flood forecasting,K-mean cluster analysis method is adopted to classify history flood.And it calibrates parameters of Xin’anjiang model in terms of flood classification.Then the flood classification is determined by calculating the distance between the flood indexes and clustering center.Finally,the exact parameters of corresponding category are chosen to forecast flood.The results show that this method not only extends the use of information,but also improves the accuracy of flood forecasting. |
| |
| Keywords: | flood forecasting K-mean cluster analysis model parameter classification |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
