最优进给方向区域性分布的复杂曲面分区加工刀轨规划方法

A large number of complex freeform surface parts exist in industries like aerospace, automotive and energy. Optimal feed directions in machining complex surfaces usually vary among different regions, which proposes urgent needs for region-based machining. Existing region-based toolpath planning methods usually construct the scalar or vector models of objective functions which refer to the nonlinear mapping from the feed direction to the objective function value. Only greedy algorithms could be used together with the discrete optimal feed direction vector field to generate region-based toolpath. As a result, only local optimal machining result could be achieved. This proposal presents a new idea to construct the rank-two tensor models for objective functions by deriving the multi-linear mapping from the feasible feed direction vector space to the objective function value domain, to establish the theoretical basis for global optimum oriented region-based machining. The conditions that whether an objective function in machining freeform surface could be modeled as tensor will be researched and then the rank-two tensor models for several typical objective functions will be constructed. Tensor based methods for surface subdivision and toolpath planning in sub-surfaces will also be proposed. This research will provide both technical and theoretical support to realise region-based optimal machining for complex freeform surfaces with regional distribution of optimal feed directions.

航空航天、汽车及能源等领域存在大量复杂曲面零件，数控加工时刀具在曲面各点的最优进给方向往往呈现区域性分布，即不同区域点的最优进给方向对应不同的走势，对分区加工提出了迫切需求。但现有的分区加工刀轨规划方法多构建目标函数的纯量或向量模型，建立曲面上一点的刀具进给方向至目标函数值的非线性映射，只能基于离散的最优进给方向矢量场采用贪婪算法生成分区加工刀轨，仅实现局部优化加工。本项目提出构建目标函数二阶张量模型的思路，建立曲面上一点的刀具可行进给方向矢量空间至目标函数值域的多重线性映射，为突破复杂曲面全局优化加工难题奠定理论基础。本项目将研究复杂曲面加工目标函数可张量化的条件，构建复杂曲面加工典型目标函数的二阶张量模型，提出基于目标函数二阶张量模型的复杂曲面加工区域划分方法及复杂曲面子加工区域加工刀轨规划方法，为最优进给方向区域性分布的复杂曲面实现分区优化加工提供理论和技术支撑。

本项目以航空航天、汽车及能源等领域复杂曲面零件优化加工为目标，突破现有基于向量方法仅能实现局部优化加工的局限，提出了基于张量的复杂曲面整体优化加工研究思路，主要成果如下：1）构建了通用回转体刀具切宽二阶张量模型，实现了曲面任意点处切宽在可行进给方向矢量空间中的全局连续度量；2）基于切宽张量场揭示了复杂曲面最优进给方向的区域性分布规律，进而提出了基于张量场三分退化点的复杂曲面整体优化分区方法，使每个区域内具有连续的最优进给方向分布，给出了分区刀轨规划方法，从而使加工过程中刀具能尽可能沿最优进给方向运动，实现了复杂曲面整体优化加工；3）基于提出的复杂曲面整体优化分区方法定义了曲面加工动态特征，并结合Function Block的事件触发机制构建了曲面加工动态特征模型，为动态制造资源环境下的复杂曲面优化加工工艺自适应决策提供了基础；4）基于本项目成果开发的原型系统已应用于某飞机型号机翼成型模具曲面零件的精加工，与商品化CAM软件相比，提高加工效率15%以上。本项目提出的基于张量的复杂曲面整体优化分区加工刀轨规划方法被CIRP Fellow英国Rajkumar Roy教授评价为“智能制造领域最引人关注的研究进展之一”，被法国工程院院士、CIRP/GARPA Fellow Alain Bernard教授评价为“能显著提高航空复杂曲面零件的加工效率”。