【摘要】 针对复杂曲面类零件加工余量分析过程中的测量数据匹配问题,提出通过初始匹配和精确匹配来实现曲面测量数据的最佳匹配。交互式的初始匹配过程决定后续算法的变量范围,精确匹配确定测量数据与曲面的最佳匹配姿态。精确匹配采用最小二乘法构造评估函数,应用边界约束BFGS方法对问题涉及的曲面匹配变换矩阵的6个未知量进行优化求解。对通过三坐标测量机获取的数据,提出了一种测头半径补偿方案。实验结果表明,该方法与遗传算法相比具有运算速度快和精度高等特点,能较好地解决复杂曲面类零件测量数据的匹配问题。更多还原

【Abstract】 An optimization method was presented for best-fitting of measurement point cloud to the complex surfaces, which consists of rough matching and accurate matching. The rough matching decided the variable range of the follow-up algorithms, whereupon the accurate matching acquird the optimal location. To cope with the accurate matching, based on a least-square method and L-BFGS-B algorithm, a new point to surface best-fitting method was put forward. For CMM data, this paper integrated a probe compensation method into the best-fitting process. The method has been compared to similar methods, such as Genetic algorithms, and is less subject to local optimization and has higher accuracy. This is demonstrated using examples at the end of the paper.更多还原