摘要
遗传算法是模型参数优化的一种有效方法。将遗传算法应用于北京市郊区不同重现期的暴雨强度与降雨历时关系式中参数的优化,并与传统回归法和优选回归法的优化效果进行了分析比较。实例计算结果表明:遗传算法用于暴雨强度公式中的参数估计精度高于传统回归法和优选回归法的参数估计精度。该方法具有直观、简便和实用特点。
Genetic Algorithm (GA) is an effective method for parameter optimization of model. Genetic algorithm was applied to optimize the parameters of storm rainfall intensity formulae of different repeat periods in Beijing suburban. Comparisons of optimum results between the genetic algorithm method and the traditional regression method as well as optimum seeking method were performed. The calculated results show that the use of GA method estimation precisions of parameters in the storm rainfall intensity formula is highter than that of traditional regression and optimum seeking methods. The method has the objectivity, simplicity and effectiveness.
出处
《高原气象》
CSCD
北大核心
2003年第6期637-639,共3页
Plateau Meteorology
基金
九五"国家重点科技攻关项目(96 911 08 03)
成都信息工程学院院管课题资助
关键词
遗传算法
暴雨强度
参数优化
Genetic algorithm
Heavy rain intensity
Parameter optimization
