论文部分内容阅读
在非线性拟合中，常采用变量替换的方法使之线性化，再按线性模型处理。本文指出按这种方法进行的参数估计不是无偏估计量，对此作者用修正回归方程系数拟合法对其参数进行优化，使之逼近无偏，提高了曲线拟合精度。通过实例计算，此法用于地下水位的动态分析效果显著。
In non-linear fitting, variable substitution is often used to make it linear, then linear model. This paper points out that the parameter estimation by this method is not unbiased estimator, and the author uses the modified regression equation coefficient fitting method to optimize its parameters so that it approximates unbiased and improves the curve fitting accuracy. By example calculation, this method is effective for dynamic analysis of groundwater level.