多项式模型错误参数更正方法探析

Exploration on Error Parameter Correction Method for Polynomial Model

在应用《湖南省森林资源调查常用数表》的过程中,洞庭湖区“三杉”多项式参数代入计算与立木材积表数据不一致,文章针对此问题的出现原因及解决方法作了细致深入的研究。通过对《湖南省森林资源调查常用数表》提供的多项式参数分析、利用Excel表进行逆向多元回归分析与最小二乘法等方式试图寻找错误,确认其可用参数与错误参数的关系,列出等式方程,并对其方程的参数求解,再用平均指标分析法作进一步比较分析,多方法求解比较,从中寻求最优解,达到改正多项式模型参数的目的。综合应用对比几种参数改进方法,测算结果表明,以平均指标分析中的胸径加权平均值求得参数为最佳,数据很吻合,可以推广应用。

When using Common Data Tables of Forest Resources Survey in Hunan Province,"Sanshan"in Dongting Lake of Hunan Province polynomial parameters computation is different from the data of the tree volume table.This paper has carried out detailed exploration on the cause and solution.Through analyzing polynomial parameters provided by Common Data Tables of Forest Resources Survey in Hunan Province,using Excel to implement inverse multiple regression analysis,as well as applying the least square method,the errors are trying to find.The relationship between available parameters and error ones is confirmed and then the equality equation is listed in a parallel way.Aiming at the equation,the parameter array is obtained and further analysis and comparison with average index analysis is made.In order to correct polynomial model parameters,the best solution is sought for by means of multi-method solution comparison.Compared with several parameter improved methods,according to the calculating result,the parameter obtained by the weighted average of breast height in average index analysis is the best,and the data is very consistent and can be applied generally.