模拟退火算法在新安江半分布水文模型中的应用  

Application of Simulated Annealing Algorithm in Xinanjiang Model

作　　者：赵维[1] 李兰[1] 武见[1] 

机构地区：[1]水资源与水电工程科学国家重点实验室武汉大学,武汉430072

出　　处：《中国农村水利水电》2008年第10期4-6,共3页China Rural Water and Hydropower

基　　金：国家自然科学基金资助项目(2003CB415206);国家重点基础研究"973"课题(2003CB415206)

摘　　要：半分布新安江模型能够反映水文的空间变异规律,比集总模型能更好的反映下垫面的影响。但是多个子流域单元的参数优选变得更加复杂,这是半分布水文模型急需解决的难题。模拟退火算法是局部搜索算法的扩展,能够跳出局部循环,并在理论上以概率1.0接近全局最优值。模拟退火算法应用在半分布新安江模型参数优选工作中,确定了参数的自动优化问题,并通过实例检验了模拟退火算法在半分布水文模型中的应用。The distributed Xinanjiang Model is able to reflect the rule of hydrology and the effect of underlying. However, too many sub basins make the optimization of parameters more complicated and this is an urgent problem that should be resolved in the distributed model. Simulated annealing algorithm has a very strong global optimization capability of jumping out of the local circle, and the overall probability of getting a nearly optimal value. This paper makes a connection between Xinanjiang Model and the SA algorithm. At last, it is proved that SA global optimization can calibrate the model globally and exactly through an example.

关 键 词：三水源新安江模型 模拟退火算法 自动优化 参数识别 

分 类 号：TV124[水利工程—水文学及水资源]