【作者】 张松；

【导师】 何改云；

【作者基本信息】 天津大学 ， 机械制造及其自动化， 2010， 硕士

【摘要】 随着现代制造业的快速发展,人们对零部件轮廓构造提出多元化要求的同时,对零部件表面质量的要求也相应提高。加工质量推动着几何量检测技术的发展。布点方法与误差评定算法对提高检测质量和效率有着直接的影响。本文主要对复杂曲面的布点策略和误差评定方法进行了研究。本文首先对复杂曲面测点分布问题进行了研究,提出了一种基于距离的曲线、曲面布点方法。根据曲线-曲面的关系,在对曲线布点方法进行研究的基础上,将其扩展来解决曲面的采样问题。根据加工要求对曲线进行初始测点分割,然后逐层求精、迭代选点,实现了曲线的自适应布点。进而应用初始网格点与基于距离的迭代选点相结合的采样方法,实现了曲面测点的自适应分布。该方法使测点能够随曲面曲率变化进行疏密分布,有效的反映曲面的拓扑信息。其次,对复杂曲面的误差评定方法进行了研究。利用最小二乘法构建误差评定的目标函数。通过对齐局部坐标系和构建点对约束条件,优化了初始条件,然后应用迭代最近点(ICP)算法,对影响曲面位移与旋转的六个参数进行优化,从而完成了复杂曲面轮廓度误差的评定工作。根据加工中的基准情况和曲面成型理论,针对大坡度以及曲率变化平缓的曲面提出了投影转换坐标、用平面度最小区域法求解的方法。最后利用仿真实验对上述算法进行了相关验证。采用基于距离的曲面布点方法,对仿真曲面进行自适应布点,并通过添加误差项来模拟真实测点,然后应用ICP法和投影法分别对其进行了误差评价。实验表明,文中方法稳定有效,具有实用性。更多还原

【Abstract】 With the rapid development of modern manufacturing industry, diversification requirements on the parts contour structure are promoted, while requirements of the surface quality are correspondingly increased. Development of geometric measurement is promoted by high processing quality. Improvement of inspection quality and efficiency is directly affected by point distribution method and error evaluation algorithm. The point distribution strategy and error evaluation on sculpture surface are mainly studied in this paper.First of all, point distribution is studied based on surface processing sampling problems and a new method of sampling on curve and surface based on distance is approved. In view of curve-surface relationship, the method of point distribution on the curve is studied and then it is extended to address the issue of surface sampling. Curve is divided by initial measurement points according to processing requirement and then adaptive point distribution is achieved by refinement and iterative search. Adaptive point distribution on surface is realized by the method on combination of initial grid points and iterative points based on distance. Point density distribution can be varied with the change of surface curvature and surface topology information is effectively reflected.Then, error evaluation method on surface is studied. The objective function is set up using least square method. The initial conditions are optimized by the alignment of local coordinate system and points constraint constructed. Profile error evaluation is completed by using iterative closest point (ICP) algorithm, while six parameters relative to surface pose are optimized. In view of reference and surface forming theory, an approach is promoted for surface with large scope and curvature changed slowly. Firstly coordinates are projected, and then the minimum zone method for flatness solving is used.Finally, the algorithms mentioned are verified by simulation experiments. The adaptive point distribution on simulation surface is realized by the surface point distribution method based on distance. And the actual measuring points are simulated by adding error term. Then error evaluation is completed by ICP and projection method, respectively. The results of experiment show that the methods mentioned are effective and practical.更多还原