| 确定天然河流纵向离散系数的有限差分单纯形法 |
| |
| 引用本文： | 薛红琴,赵尘,刘晓东,顾莉.确定天然河流纵向离散系数的有限差分单纯形法[J].解放军理工大学学报,2012,0(2):214-218. |
| |
| 作者姓名： | 薛红琴 赵尘 刘晓东 顾莉 |
| |
| 作者单位： | 1.南京林业大学 土木工程学院,江苏 南京 210037; 2.河海大学 浅水湖泊综合治理与资源开发教育部重点实验室,
江苏 南京 210098; 3. 河海大学 环境学院,江苏 南京 210098 |
| |
| 基金项目： | 国家973计划资助项目（2008CB418202）；国家水体污染控制与治理科技重大专项资助项目(2009ZX07103 006)；中央高校基本科研业务费专项资金资助项目（2009B17814）. |
| |
| 摘 要： | 针对在确定纵向离散系数的示踪试验中解析法难以适用于天然非均匀河流的不足，提出了采用有限差分法结合Nelder-Mead单纯形法的参数识别算法(FDM-NMS)来反演天然河流的水质参数。以纵向离散系数反演为例，利用该算法重点探讨了初值选取、观测噪声、离散河段数等因素对参数识别结果的影响，结合恒定流和非恒定流2个算例验证了该方法的可靠性。计算结果表明，采用FDM-NMS算法当噪声水平≤10%、离散河段数≤10时能给出较好的参数识别结果，算法具有良好的抗噪性。
|
| 关 键 词： | 参数识别 水质模型 有限差分 单纯形法 |
| 收稿时间： | 2010-11-02 |
| 修稿时间： | 2010-11-02． |
Finite difference method simplex method for determination of
logitudinal dispersion coefficient in natural river |
| |
| XUE Hong-qin,ZHAO Chen,LIU Xiao-dong and GU Li.Finite difference method simplex method for determination of
logitudinal dispersion coefficient in natural river[J].Journal of PLA University of Science and Technology(Natural Science Edition),2012,0(2):214-218. |
| |
| Authors: | XUE Hong-qin ZHAO Chen LIU Xiao-dong and GU Li |
| |
| Institution: | 1. College of Civil Engineering, Nanjing Forestry University, Nanjing 210037, China;
2. Ministry of Education Key Laboratory of Integrated Regulation and Resource Development on Shallow Lakes,
Hohai University, Nanjing 210098, China; 3. College of Environment, Hohai University, Nanjing 210098, China |
| |
| Abstract: | Given the difficulties of the analytical method which can't be appliced to the non uniform river in tracertest for determination of longitudinal dispersion coefficient, a parameters identification method based on finite difference method nelder mead simplex (FDM NMS) was proposed. Taking longitudinal dispersion coefficient as example, the factors such as the different level of the observation noise, the discreted stream segment number and the initial value selected which affects the results of the parameters estimation were discussed in detail by the given algorithm. The method was validated by two numerical examples: parameters identification in steady and unsteady flow. The computational results indicate that FDM NMS algorithm can give good identification precision results and shows good noise proof for the water quality model with the noise level ≤10% and discreted stream segment number ≤10. |
| |
| Keywords: | parameters identification water quality model FDM(finite difference method) simplex |
|
| 点击此处可从《解放军理工大学学报》浏览原始摘要信息 |
| 点击此处可从《解放军理工大学学报》下载免费的PDF全文 |
