动态规划算法在河道糙率反演中的应用
详细信息    查看全文 | 推荐本文 |
  • 英文篇名:The Application of Dynamic Programming in River Roughness Inversion
  • 作者:于显亮 ; 彭杨 ; 吴志毅
  • 英文作者:YU Xian-liang;PENG Yang;WU Zhi-yi;School of Renewable Energy,North China Electric Power University;
  • 关键词:河道糙率 ; 反演模型 ; 动态规划法 ; 一维恒定流
  • 英文关键词:river roughness;;inverse model;;dynamic programming;;one-dimensional steady flow
  • 中文刊名:ZNSD
  • 英文刊名:China Rural Water and Hydropower
  • 机构:华北电力大学可再生能源学院;
  • 出版日期:2017-05-15
  • 出版单位:中国农村水利水电
  • 年:2017
  • 期:No.415
  • 基金:长江科学院开放研究基金资助项目(CKWV2015201/KY);; 国家自然科学基金面上项目(51679088)
  • 语种:中文;
  • 页:ZNSD201705020
  • 页数:5
  • CN:05
  • ISSN:42-1419/TV
  • 分类号:92-95+100
摘要
河道糙率是水力计算的关键参数,在实际工程中一般根据实测资料建立经验公式估算或采用试错法来确定糙率,不仅工作量大且具有较大的经验性和任意性。以河道一维恒定流计算为例,将糙率视为流量的函数,以水位误差平方和最小为目标,基于动态规划法建立了河道糙率反演模型,将河道不同流量级下的糙率反演问题转化为一个多阶段的糙率优化问题,并将其应用于向家坝库区河道恒定流计算的糙率反演中。计算结果表明,建立的模型具有良好的拟合效果,且糙率区间内离散点数越多,拟合的效果越理想,当离散点数为56、计算精度为0.002时得到的水位最大平均误差不超过15 cm。该方法能有效避免手工调试的不确定性,为实现糙率精确率定与提高水力仿真模拟精度提供技术支持。
        River roughness is an important parameter of hydraulic calculation. In practical projects,the value of roughness is usually estimated by empirical formulae or through trial and error method according to the measured data. In this paper,taking one-dimensional steady flow as an example,an optimized inverse model based on dynamic programming algorithm is developed to estimate Manning's roughness parameter,in which the roughness is regarded as a function of discharge and the minimum square sum of calculation error of water stage is set as the target function. Thus the inverse problem of roughness under different flow levels is transformed into a multi-stage roughness optimization problem. The developed model is applied to inverse the roughness parameters of river course in Xiangjiaba Reservoir. Results show that the model has a good inversion effect and the more discrete points are,the better the imitative effect is. When the discrete points are 56 and calculation precision is 0.002,the maximum average error is less than 15 cm. This method can effectively avoid the uncertainty of the manual debugging,and provides a technical support to achieve precise calibration of roughness and improve the hydraulic simulation accuracy.
引文
[1]Becker L,Yeh W W G.Identification of parameters in unsteady open channel flows[J].Water Resources Research,1972,8(4):956-965.
    [2]Wasantha Lal A M.Calibration of riverbed roughness[J].Journal of Hydraulic Engineering,1995,121(9):664-671.
    [3]金忠青,韩龙喜,张健.复杂河网的水力计算及参数反问题[J].水动力学研究与进展(A辑),1998,13(3):280-285.
    [4]程伟平,毛根海.基于带参数的卡尔曼滤波的河道糙率动态反演研究[J].水力发电学报,2005,24(2):123-127.
    [5]张潮.基于机器学习的河网糙率反演[D].杭州:浙江大学,2008.
    [6]刘志贤.基于遗传算法的河网糙率反分析研究[D].杭州:浙江大学,2008.
    [7]李丽,王加虎,王建群,等.自适应随机搜索算法在河网数学模型糙率反演中的应用[J].水利水电科技进展,2011,31(5):64-67.
    [8]齐鄂荣,罗昌.库区河道非恒定流糙率的选取及特性[J].武汉大学学报(工学版),2003,36(2):1-5.
NGLC 2004-2010.National Geological Library of China All Rights Reserved.
Add:29 Xueyuan Rd,Haidian District,Beijing,PRC. Mail Add: 8324 mailbox 100083
For exchange or info please contact us via email.