摘要
感潮水闸流量的准确计算对于河网地区水闸引排水效益的分析和水闸综合管理体系的建立具有非常重要的意义。对感潮水闸的水力特性进行了详细分析,认为感潮水闸具有瞬时性和非线性等水力特点,提出采用人工神经网络理论建立其过流量的计算模型。建立了三种计算模式,应用浦东新区东沟水闸资料对不同模式进行了训练、测试和比较,推荐以水闸内外河水位、闸门开启度和上一时刻流量作为神经网络输入的计算模型为最终感潮水闸流量计算模型。研究表明,人工神经网络可以较好地隐含识别感潮水闸的多种出流类型,并具有较强的泛化和容错能力,从而为感潮河网地区水闸流量的计算提供了一种新的解决途径。
Correct calculation of tidal sluice discharge is important for benefit analysis of water allocation of sluices in river networks and for constructing the integrated management system of sluices. It is concluded that the tidal sluice has nonlinear and transient hydraulic characteristics after a detailed analysis of the tidal sluice＇ s observed data. Then the calculation model of the tidal sluice discharge is presented on the basis of artificial neural networks（ANN） theory. Three calculation patterns of the model are established, trained, tested, and compared with Donggou sluice＇ s data in Pudong New Area. An ideal pattern is finally recommended as the calculation model of tidal sluice discharge, which has the inputs of sluice＇ s inside and outside river stage, gate opening and discharge at previous time. This paper shows that ANN can be used to implicitly recognize tidal sluice＇s outflow types, has good generalization and tolerance capabilities, and provides a new way for settling the discharge calculation of the tidal sluice in the river-network region.
出处
《海洋工程》
CSCD
北大核心
2007年第3期109-114,共6页
The Ocean Engineering
基金
杭州市科技发展计划资助项目(20051331B06)
浦东环保市容局科研资助项目(PHK2006014)
关键词
水闸
流量
人工神经网络
感潮河网
sluice
discharge
artificial neural networks
tidal river networks
