基于遗传算法的皮尔逊Ⅲ型曲线参数估计

Pearson Type Ⅲ Curve Parameter Estimations Based on Genetic Algorithm

皮尔逊Ⅲ型频率曲线参数估计时,常用方法有矩法、权函数法和适线法等,这些方法各有缺陷。根据遗传算法全局寻优的原理与特点,采用遗传算法计算皮尔逊Ⅲ型频率曲线的三个参数,并和其他方法进行了对比验证。结果表明,遗传算法得到的频率曲线与实测资料误差小,拟合精度高,是一种较好的计算方法,值得推广应用。

Moment method, weighted function method, curve-fitting method and so on, which have some defects, are used to estimate the parameters of Pearson type-Ⅲ distribution. According to the characteristics of genetic algorithm of overall optimization, genetic algorithm is used to estimate the parameter of Pearson Type-Ⅲ curve in this paper, and is compared with other methods. The result indicated that the Pearson Type-Ⅲ curve gained from genetic algorithm approaches the data, and the fitting precision is high. Consequently the method is worth popularizing.