遗传算法应用于多元非线性回归模型求参的研究  

The study on determination of multivariate nolinear regression coefficent about gentic algorithm

作　　者：霍倩[1] 李书全[1] 王文元[1] 赵胜利[1] 赵春生[1] 

机构地区：[1]河北农业大学城建学院,河北保定071001

出　　处：《河北农业大学学报》2002年第2期107-110,共4页Journal of Hebei Agricultural University

摘　　要：传统的非线性回归模型求参方法存在精度不高、限制条件繁琐等问题。本文提出了遗传算法这一有效算法 ,由 2个算例给予验证 ,并与线性最小二乘法和Gauss Newton迭代法的结果作了对比 。The old methods to determine multivariate nolinear regression coefficient have low precision and many restricted application conditions Genetic algorithm as one efficient method is suggested in this paper, and is also indicated by two examples comparing with linear least square method and Gauss Newton method The results indicate that genetic algorithm is a more universal way

关 键 词：多元非线性回归模型 遗传算法 模型参数 求参方法 

分 类 号：O212.1[理学—概率论与数理统计] O242.23[理学—数学]