【作者】 李剑；

【导师】 陈子辰； 王文；

【作者基本信息】 浙江大学 ， 机械制造及其自动化， 2002， 博士

【摘要】 信息社会中制造业的产品具有信息模型和实物两种不同范畴的表现形式，而产品生产中最具创新和价值的部分被包含在产品信息模型中，对于在现代工业中得到广泛应用的自由曲面零件尤为如此。所以，对自由曲面零件进行三维几何形状测量、数学建模和数控加工及产品表面质量评定，在现代制造业中具有重要的现实意义。本文在综述国内外关于自由曲面零件的数控测量方法、测量规划、曲面造型、测量造型加工(3M)一体化技术研究现状及发展趋势的基础上，综合利用激光测量、精密机械、光电子、CAD／CAM和自动控制等学科技术，研制开发了基于激光点光源测头的非接触式数控测量原型系统和基于现有数控系统的自由曲面非接触测量加工一体化系统。同时对所涉及的若干自由曲面零件测量、重建、加工和检测中的相关问题进行了深入研究。 1． 自由曲面的高速、精确、自适应测量是自由曲面数字化制造需要解决的首要问题。为此，首先分析了激光点光源测头的测量原理和自由曲面物体表面三维坐标的拾取原理；提出了一种曲面零件的多模分段自适应测量规划方法和矩形自适应测量规划方法，并进行了仿真研究，使得在满足测量精度要求的前提下，测点的数目尽可能得少；提出了测量区域边界的自动跟随测量方法，以便于测量路径的优化和测点数据重建；将扫描优角顶点的概念和单调剖分的思想用于自由曲面测量路径的优化中，并通过实例验证该方法的有效性。 2． 对非接触测量得到的大量的曲面测点数据进行滤噪、压缩、分片和光顺等预处理。提出了一种模糊加权均值滤波方法，该方法利用模糊隶属度思想，通过迭代优化权值，以剔除测点数据噪声，相对于均值滤波，能更有效地剔除测量数据中各种噪声信息；提出了一种数据压缩算法。在一定的精度条件下，该方法可对包含大量冗余的测量数据进行压缩处理；对测量数据进行分片处理，拾取测量数据中的阶跃边界点、褶皱边界点和光滑边界点并优化测点数据，从而便于对测量数据进行NURBS(Non-Uniform Rational B-Spline)重建和加工路径生成等后续处理。 3． 测点数据的NURBS重建。对经预处理后的测点数据，通过目标函数的修正和迭代及最优控制点数目的求取，进行测点数据的非均匀有理B样条曲线拟合，对得到的有理B样条曲线进行一致化和蒙面处理，得到曲面的NURBS表示，再将其转换成标准的IGES文件形式，以便于通过CAD／CAM软件对待测曲面做进一步的处理。仿真和实验结果验证了该方法的可行性。此外，还研究了重建曲面的精度评定和曲面拼接方法。 4． 研究由测点数据对自由曲面零件直接进行加工路径规划的方法。先对测点数据进行局部参数曲面逼近，提取其法矢和主曲率等关键几何信息；基于最小曲率半径的刀具，对各种类型自由曲面沿切削方向和切削进给方向统一规划刀具路径；为了进一步提高加工效率，选用不同半径的刀具和求取干涉区域，优化刀具路径以高效地实现自由曲面零件的分片加工，仿真结果验证了该方法的有效性。 5． 针对复杂曲面零件检测过程中难以准确定位和测量效率低的问题，提出自由曲面零件的寻位自适应检测新方法。对该方法所涉及的曲面零件的自动寻位、测点的自适应生成和基于寻位信息的位姿自适应检测控制等关键问题，给出了可行的解决办法。在此基础上，开发了曲面零件寻位自适应检测软件。理论分析与仿真实验结果表明，所提出的方法可行，为复杂自由曲面零件的无定位检测开辟了一条新的有效途径。 6． 研制自由曲面激光测量原型系统，分析系统的硬软件组成、测量原理和各模块的功能；为实现制造自动化过程的低成本，研制开发了一套基于现有三维数控铣床系统的通用激光测量加工系统，分析了其实现原理和硬软件组成。通过实例对前面各章理论进行验证，实现了自由曲面零件的测量、重建和加工路径的生成，为自由曲面零件的数字化制造奠定了基 浙江人学博土学位论文：基于激光测量的自由曲面数宇制造基础技术研究 一 6th。 最后，总结本论文研究工作的主要结论，并对后续研究工作作了展望。更多还原

【Abstract】 The manufacturing products in the current information age have such two kinds of different representations as information model and real product, while the most innovative and valuable part of the manufacturing is included in the former one. This is the case especially for the freeform surface parts which are widely used in modern industry. Thus such research work as the measuring, modeling, manufacturing and quality assessment of freeform surface parts is of great importance in modern manufacturing industry. In this dissertation, based on the survey of the present research situation and development trends at home and abroad of the numerical control measurement methods, measurement layout, surface modeling, the integration technology of measuring, modeling and manufacturing of freeform surface parts, by absorbing techniques such as computer vision, precision mechanism, photoelectron, CAD/CAM and automatic control, etc., a non-contact numerical control measurement prototype system using probe of laser spot-source and an integration system of non-contact measurement and manufacturing of freeform surface based on the existing numerical control milling machine were developed, meanwhile the relevant problems on the measuring, modeling, manufacturing and inspection of the freeform surface parts are profoundly studied.The main contributions of this dissertation are briefly presented as follows:1. The rapid, precise and adaptive measurement of freeform surface is the primary problem to solve in its digital manufacturing. Thus, the measurement principle of the probe of laser spot-source and the method to collect the coordinates of freeform surface were firstly analyzed. Then, a multi-pattern adaptive measurement planning method in one dimension and a rectangle -adaptive measurement planning method in two dimensions were put forward and simulated so as to minimize the number of measurement points under certain measurement accuracy. An automatic tracing measurement method of the boundary of measurement area was proposed thus to optimize the measurement path and reconstruct the measurement points. The concept of scanning reflex vertex and monotone chains were used in the optimization of the measurement path of freeform surface, moreover, the effectiveness of the method was verified by example.2. To filter noise in the huge amounts of measurement points by non-contact measurement method and compress, segment and smooth the measurement points, a fuzzy weighted average filter method using fuzzy membership idea to optimize the weights by iteration method was proposed. Compared with average filter method, the method can eliminate the noise information in the measurement points more effectively. A data compression method was proposed to compress the measurement points that include huge redundancy under certain accuracy. Then segmentation and smoothing method for the measurement points to identify the jump edge, crease edge and smooth edge and optimize the measurement points were conducted so as to reconstruct the surface and generate the manufacturing path conveniently.3. NURBS reconstruction of measurement points, the measurement data after pre-filter were fitted with non-uniform rational B-spline curves with modification of the objective function, iteration method and optimization of the number of control points. The B-spline curves obtained were compromised and skinned to acquire the NURBS representation of the surface and then were transformed into standard IGES format to further dispose the measured surface with commercial CAD/CAM software. The results of computer simulation and experiment verified theeffectiveness of the method. In addition, the surface merging method and accuracy assessment method of the reconstructed surface were studied.4. The method to plan the manufacturing path directly from the measurement points was studied. The key geometry information such as normal vector and main curvature etc. were firstly calculated by local approximation of measurement points wi更多还原