作者在最小二乘法的基础上,提出分段直线拟合算法来代替多项式曲线拟合,从而达到简化数学模型的建立和减少计算量的目的。
The author proposes a sectional line fitting to re - place the polynomial curve fitting, which results in simplifying establishment of mathematical model and reducing the amount of calculation work.
同时,拟合方法采用目前非常成熟的多项式最小二乘拟合。从而不仅能控制拟合函数的复杂性,而且使分段拟合的补偿区间数量达到最少,极大地降低了计算成本。
The paper adopts the method of multinomial least squares as fitting to make the fitting function be simple and the amount of compensation interval be least, reduces the calculation cost.
计算结果表明,用分段多项式最小二乘法拟合计算M F D,可以取得满意的计算精度。
The error sources and their influences to the measurement results are also discussed. It has been shown, that the satisfied results can be achieved by using least square fitting method.
应用推荐