摘要
为强化黄河源区水资源管理及生态环境规划等问题,采用HBV模型、新安江模型对其流量过程进行模拟。HBV模型中共有15个参数,参数率定采用蒙特卡罗法(MC)和遗传算法(GA),发现敏感参数有5个,且HBV模型中应用遗传算法计算得到的结果与实测过程更加接近。新安江模型共有17个参数,参数率定采用了蒙特卡罗法(MC)、拉丁超立方分层抽样(LHS)、次数多的蒙特卡罗法,发现敏感参数共有10个,大多数时候拉丁超立方方法率定结果与实测值更接近。两个模型应用结果对比发现,HBV模型对枯季的模拟流量偏低,新安江模型对枯水年的模拟比HBV模型好,且对枯水年的模拟均处于可接受范围之内。
In order to strengthen water resources management and ecological environment planning,HBV model and Xin＇anjiang model were used to simulate the flow process of source region of the Yellow River.There are 15 parameters in HBV model.Monte Carlo and genetic algorithm were adopted to calibrate the parameters,and 5sensitive parameters were found.What＇s more,the results by genetic algorithm in HBV model are more consist with the measured values.Xin＇anjiang model has 17 parameters.Monte Carlo and Latin hypercube sampling methods were adopted to calibrate the parameters.It found out that there are 10 sensitive parameters.Most of the time,the results of Latin hypercube sampling method basically agrees well with the measured values.Comparison of these two models,the simulation of dry season＇s runoff with the HBV model is lower;instead,Xin＇anjiang model is better in simulation of dry years,which is in the acceptable range.
出处
《水电能源科学》
北大核心
2016年第12期41-45,14,共6页
Water Resources and Power
基金
国家自然科学基金项目(41371051)
关键词
HBV模型
新安江模型
黄河源区
蒙特卡罗法
遗传算法
拉丁超立方分层抽样
HBV model
Xin＇anjiang model
source region of the Yellow River
Monte Carlo method
genetic algo rithm
Latin hypercube sampling
