闸门对河道污染物影响的模拟研究  

Modeling impaction of gate on contamination transfer in stream channel by ANN

作　　者：阮燕云[1] 张翔[1] 夏军[2] 张永勇[2] 张洪斌[1] 

机构地区：[1]武汉大学水资源与水电工程科学国家重点实验室,湖北武汉430072 [2]中国科学院地理科学与资源研究所陆地水循环及地表过程重点实验室,北京100101

出　　处：《武汉大学学报（工学版）》2009年第5期673-676,共4页Engineering Journal of Wuhan University

基　　金：国家自然科学基金国际合作重大项目资助(编号:40721140020)

摘　　要：通过明渠水槽模拟的闸门调控实验,初步认识了闸门影响下污染物迁移转化波形分布规律.实验结果表明,通过合理调度,闸门在一定开度下可以起到既保证上游污染物通过闸门、减少闸前污染物质的聚集,又可降低闸下河道沿程的污染物浓度分布的作用,为进一步在实践中确定合理的考虑污染控制的闸坝调控方案提供了实验依据.为进一步认识闸门调控下污染物迁移规律,本文采用实验数据,建立了模拟河道流量、闸上游水位、闸前污染物浓度、闸门高度与闸下游各断面的污染物浓度之间复杂非线性关系的BP神经网络模型,网络学习和检验都取得了较高的精度,表明了人工神经网络在闸门对河道污染物影响的模拟与预测中的可行性和实用性.Through the gate operation experiment in open channel, we get a cognition about the phenomena of concentration wave of contamination under the influence of gate. The experimental results show that through optimal operation, the gate with a certain opening height can not only ensure the upper contamination going through the gate, reducing the accumulation of contamination in front of the gate, but also mitigate the contamination concentration distribution along the downstream of the gate. For the further practices, these provide experimental evidences to determine the optimal pollution control of gate operation schemes. In order to understand the contamination＇s transfer law more deeply, this paper uses the experimental data to establish an artificial neural networks（ANN） BP model, which simulates the complicated nonlinear relationships between the streamflow, water level at the upstream of gate, contamination concentration in front of the gate, gate opening height and contamination concentration at the downstream of the gate. The high accuracy of training and testing are obtained so as to show that the artificial neural networks is feasible and practical in the simulation and prediction for the impaction of gate operation on the contamination transfer in stream.

关 键 词：闸门调控 实验 污染物迁移规律 人工神经网络 

分 类 号：X522[环境科学与工程—环境工程]