摘要
基于贝叶斯原理,将马尔科夫链蒙特卡罗法(MCMC)与地下水数值模拟软件MODFLOW进行耦合,用于解决水文地质模型建立过程中的参数优选问题。通过调节似然函数的权重因子,来提高程序敏感性,从而大大提高参数优选的精度和效率。实例研究表明,该方法适用于水文地质问题中的参数优选问题,并体现出优异的全局寻优特性和较高的优选效率。
Based on Bayesian theory, the Markov chain Monte Karlo (MCMC) method, coupled with MODFLOW groundwater nu- merical simulation software, is adopted to solve the parameters optimization problem in hydrogeological model. Through adjusting the weighting factor of likelihood function, the sensitivity is improved, thus the accuracy and efficiency of parameters optimization is greatly improved. Case study shows that the approach is suitable for the parameters optimization problem in the hydrological geolo- gy, and reflects the excellent global optimization characteristics and high optimization efficiency.
出处
《节水灌溉》
北大核心
2013年第12期33-36,39,共5页
Water Saving Irrigation
基金
吉林省重点科技攻关项目(20100452)
"十一五"国家科技支撑计划项目(2008BAB42B07)
吉林省科技发展项目(20080543)