摘要
本文提出一种匹配拟合曲面和已知长短轴的标准椭球曲率的匹配算法来修正拟合曲面方程,以改善拟合精度。该算法首先合理地选取反映被测小面形特性的特征点,用最小二乘法拟合出椭球方程,然后利用刚性曲面在空间旋转平移后曲率不变的特性,在给定长短轴且正置的椭球表面上搜索与拟合面形上的特征点的曲率相同的点,进而得到和测量小面形的高精度匹配。实际生产应用的结果表明该方法是十分有效的。
In this paper,a new method of curvature matching algorithm improving fitting precision of fitting rotational ellipsoid equation is proposed.Firstly,the algorithm needs to select some characteristic points indicating property of tested surface.Secondly,a fitting ellipsoid equation is given by least square algorithm,and then utilizes curvature has an invariant property with respect to the shift and rotation of the rigid surface,and searching points on the surface of the ellipsoid which known major axis length and minor axis length,which their curvature is equal to the characteristic points on fitting surface,and then gained high precision matching of a small tested surface.The new method is successfully applied into making shelled antenna.It is shown the new method is an effective one in the field of rotational ellipsoid fitting.
出处
《计算机应用与软件》
CSCD
北大核心
2005年第6期95-96,118,共3页
Computer Applications and Software
