摘要
基于暴雨推求中小流域桥梁设计流量时,设计流量受地理参数、暴雨参数等众多因素的影响,计算公式难以显式表达的情况,采用H-C-N法计算中小流域桥梁设计流量:用层次分析法(Analytic hierarchy process)对影响流量的因素进行分析并筛选出主要因素;用模糊聚类法(Fuzzy clustering)从已知水文站数据集中选取合理的训练样本;运用神经网络方法(Neutral network)建立中小流域桥梁设计流量的隐函数关系式,从而计算出中小流域桥梁的设计流量。用VB.NET编写程序编写H-C-N法计算程序,计算了10座中小流域桥梁的设计流量。结果表明,与广东省法和四院法相比,H-C-N法精度较高,是一种较为科学和实用的算法。
The bridge design water flow under rain- storm is influenced by many factors. Therefore,it's very difficult to find out the calculation formula. In order to work out the bridge design water flow of middle and small river basin, the H - C - N(analytic hierarchy process - fuzzy clustering - neutral network)method was introduced. Main factors influencing the design water flow were analyzed by analytic hierarchy process(AHP), and the training samples were selected by fuzzy clustering method. The training samples were chosen from the forgone hydrologic data. Consequently, a BP Neutral Network model was established to characterize the functional expression. In this way, the design water flow was worked out. The arithmetic was realized by VB. NET. Design water flow of ten bridges was calculated. Compared with Guangdong method and the method of the forth design institute, the precision of H - C - N method is more satisfactory. According to the results, the H - C - N method is scientific and useful.
出处
《铁道科学与工程学报》
CAS
CSCD
北大核心
2008年第2期42-45,共4页
Journal of Railway Science and Engineering
基金
广东省交通集团科技研究开发项目(KJXM-010)
关键词
中小流域
设计流量
H—C—N法
神经网络
层次分析
模糊聚类
middle and small river basin
design water flow
H - C - N method
neutral network
analytic hierarchy process
fuzzy clustering