摘要
利用水头实测资料,以裂隙组的渗透系数比例因子为待反演的参数向量,基于浮点编码的加速遗传算法,提出裂隙渗流参数的进化反演模型,并通过典型坝基渗流算例进行验证。文中同时给出了传统单纯形算法在不同初始值条件下的反演成果。计算结果表明,本文方法可以在仅给出各参数取值范围的条件下,有效地求得非均质各向异性裂隙岩体渗流参数,且克服了传统优化搜索算法严重依赖于初值,且当待反演参数数目较多时,寻优过程极易陷入局部最优值,甚至搜索不收敛等缺陷。
A model for back analysis of seepage parameters in fractured rock mass is proposed.On the basis of floating point accelerated genetic algorithm,the seepage coefficient scale genes of different fracture groups are selected as the objective parameters and the measured water head data are to be used in the calculation.The validity of the method is verified by the calculation result of the foundation seepage of a typical gravity dam.It shows that this method is effective in identifying the seepage parameters of fractured rock mass if the ranges of the parameters are given.It not only accelerates the calculation but also overcomes the shortcomings of traditional method in optimization.
出处
《水利学报》
EI
CSCD
北大核心
2003年第2期55-60,共6页
Journal of Hydraulic Engineering
基金
国家自然科学基金资助项目(5990902)
关键词
裂隙岩体
渗流参数
反演
遗传算法
单纯形算法
fractured rock mass
seepage parameter
back analysis
genetic algorithm
