摘要
传统差分进化算法在优选水文模型参数时容易出现"早熟收敛"问题,基于马尔可夫链蒙特卡罗方法的差分进化算法——DREAM算法,对嘉陵江流域降雨径流模型的参数优选问题进行了分析。结果发现,DREAM算法融合了自适应Metropolis方法的优点,能有效克服"早熟收敛"问题,适用于推求先验信息较少的复杂水文模型参数后验分布。
Premature convergence problem exists in the parameter optimization of hydrological model using the traditional differential evolution.In this paper,differential evolution adaptive metropolis algorithm（DREAM）was proposed,which combines the advantages of differential evolution algorithm and Markov Chain Monte Carlo（MCMC）sampler,and applied in the parameter optimization of CMD-3PAR hydrologic model in the Jialing River Basin.The results showed that DREAM has the advantages of self-adaptive Metropolis method,can effectively overcome the problem of premature convergence,and is capable to infer the posterior distribution of model parameters which is lack of prior information.
出处
《南水北调与水利科技》
CAS
CSCD
北大核心
2015年第2期202-205,共4页
South-to-North Water Transfers and Water Science & Technology
基金
国家自然科学基金(50809058)
关键词
差分进化算法
MCMC方法
参数优选
DREAM算法
降雨径流模型
differential evolution algorithm
Markov Chain Monte Carlo method
parameter optimization
Differential Evolution Adaptive Metropolis
rainfall-runoff model
