[1]HU Caihong,WANG Yanju,WU Zening.Application of Support Vector Machine Based on Clusteringin Flood Foreeast[J].Journal of Zhengzhou University (Engineering Science),2009,30(04):123-127.
Copy
Journal of Zhengzhou University (Engineering Science)[ISSN
1671-6833/CN
41-1339/T] Volume:
30卷
Number of periods:
2009年04期
Page number:
123-127
Column:
Public date:
1900-01-01
- Title:
-
Application of Support Vector Machine Based on Clusteringin Flood Foreeast
- Author(s):
-
HU Caihong; WANG Yanju; WU Zening
-
School of Water Conservancy and Environment Engineering,Zhengzhou University,Zhengzhou 450001,China
-
- Keywords:
-
system clustering; support vector machine; flood forecasting
- CLC:
-
-
- DOI:
-
-
- Abstract:
-
The difficulty of the runoff simulation has been increased due to the specificity of the semi—arid re.gion,which is still a hot and difficult topic in current study.The detailed information is hard to obtain becauseof the complicated process of the runoff,SO it is difficult to forecast flood,especially the flood peak forecast.The different characteristics of flood can not be reflected completely if all samples were used to calibrate theparameter of the model.Thus the support vector machine method based on clustering is used.And the LanRiver basin in the seni—arid region is taken as an example to be simulated and tested.The results haveshown that most efficient eoemcient iS beyond 85%.and the modulus of the relatively average error is mostlyless than 1.5%.Additionally,the peak flow modulus of the relatively average error is less than 1 5%。particu.1arly in the flood of large peak flow.the relatively average error is less than 1%.The qualification rate of thepeak flow and the peak time difference is 1 00%.