Freeform Surface Profile Error Evaluation Based on Gray Code Genetic Algorithm and Segmentation of Spherical Approximation Method

摘要

In this paper, aiming at the existing problems of free surface profile error evaluation, the cubic non-uniform rational B-spline is applied to describe the surface. In order to make the free surface more close to the collection of data points, the improved genetic algorithm optimization is adopted to reconstruct the free-form surface, method of segmentation spherical approximation is employed to calculate the shortest distance between measuring points and the surface. The control vertices is applied by curved surface fitting to calculate the shortest distance for the first time, then split spherical approximation method is utilized to calculate the shortest distance between the surface and measuring points. Through example validation, spherical approximation method is exploited to assess error of the optimized and reconstructed surface, which has the characteristics of high precision and reliable algorithm.