自适应复杂曲面形变的点云数据高效数控加工方法研究

Cloud of points based directly high efficiency machining of complex surface, which entirely eliminates the complicated modeling process from data points to complex parametric surface, is an effective way to shorten design and manufacturing cycle time and reduce manufacturing costs. This projection is closely focused on the issue of high efficiency NC machining of cloud of points of deformed complex surface. By comparison between different representation models of complex surface, its shape change region can be accurately identified, and the original tool paths at this region, or the generated tool paths on the nominal design surface or on the base surface are used as the basic paths; Parameterization mapping technology of mesh or cloud of points is introduced and used to construct the deformation vector field between the basic path points and their target points in the measured data; On basis of this, theoretical model for the basic path deformation is established and the non-rigid transformation is solved using a general fashion of alternate iteration strategy, in which the operations of rigid localization and free-form deformation are alternately performed to make the basic path gradually adapt to the shape changes of complex surface; Then, according to the predicted results of the machined surface topography, the path interval and path distribution are dynamically adjusted so that the machining precision of deformed tool paths can be accurately controlled. Successful implementation of this projection will provide some novel ideas and technical supports for the high efficiency machining of complex surface.

越过点云数据到复杂参数曲面的繁琐构造过程，直接基于点云实现复杂曲面零件的高效加工是缩短产品设计制造周期、降低制造企业生产成本的有效途径。本项目紧紧围绕形变复杂曲面点云数据的高效数控加工展开研究。拟通过复杂曲面不同描述模型间的比对，实现对形变区域的准确定位，以相应的原始刀轨、名义设计曲面或简单逼近基面为基础，构造高效的基本加工路径；引入网格或点云的参数映射技术，建立能保证基本加工路径点与变形目标点间正确对应关系的变形矢量场；依此建立基本加工路径自适应变形的理论模型，给出刚性定位与自由变形轮换迭代的非刚体变换求解策略，实现对加工路径位置调整和形状改变的灵活操控，使其逐渐适应复杂曲面形状的变化；进而根据对加工表面形貌的准确预测结果，完成走刀行距及路径分布的动态调整，实现对变形轨迹加工精度的精确控制。最终，通过本项目的成功实施，为复杂曲面的高效数控加工提供新的思路和技术支持。

越过点云数据到复杂参数曲面的繁琐构造过程，直接基于点云实现复杂曲面零件的高效加工是缩短产品设计制造周期、降低制造企业生产成本的有效途径。本项目紧紧围绕形变复杂曲面点云数据的高效数控加工展开研究。在完成点云数据与零件设计模型最优匹配的基础上，给出了复杂曲面零件形变区域的快速识别方法；针对所获得的形变区域数据或测量数据，发展出基于点云基面导向曲线迭代投影的无干涉数控加工路径设计方法，实现了直接面向点云的无干涉高效数控加工；以柱面映射为基础，引入离散曲面上径向曲线的概念，通过径向曲线间的螺线线性插值，提出了零/非零亏离散曲面的螺旋抑振加工方法；进而提出曲面离散点最优走刀方向到参数域的映射方法及发生走刀矢量场的流函数构造方法，给出并证明了多片曲面间走刀矢量场的光滑拼接条件，形成了兼顾离散点局部走刀方向与轨迹线族整体表达的高效加工路径设计方法；随后提出了新的几何/运动/驱动物理特性约束下的五轴加工刀具姿态优化数学模型，给出计算高效的变量分治求解策略，解决了传统Sturz法及四元数位姿优化法引起的高速加工运动速度突变问题，可缩减高达46.7%的机床旋转轴角加速度波动；从切削运动学及微元切削机理出发，建立了变进给/变倾角等实际加工工况下的五轴数控加工表面微观形貌预测模型，解决了沿任意拓扑加工路径、考虑刀轴矢量及进给速度动态变化的五轴切削过程建模及表面形貌的精确预测问题；进行了相应的模拟仿真和实际加工实验，实现了直接面向点云数据的复杂曲面高效数控加工，达到了项目预期研究目标。以第一及通讯在International Journal of Mechanical Sciences (IF: 4.134) 、Robotics and Computer-Integrated Manufacturing (IF: 4.339)、Computer-Aided Design (IF: 3.049)、美国机械工程学会（ASME）权威期刊Journal of Manufacturing Science and Engineering 和Journal of Computing and Information Science in Engineering等制造领域重要学术期刊和国际会议上发表论文13篇，授权国家发明专利5项，指导博士生2人、硕士研究生9人（毕业7人）。