摘要
针对参数曲面迭代求交在退化场景下的数值不稳定问题,提出一种融合几何特征的鲁棒性增强方法.采用凸包辅助重启点技术重构振荡迭代点,基于控制凸包几何中心计算替代点打破无限循环;设计振荡抑制矢量分解技术分解迭代矢径,通过抑制切向分量消除振荡;结合参数域仿射不变量技术,利用区域分析和雅可比矩阵反算机制稳定切触区域计算.实验结果表明:在贝塞尔曲面测试中,提出的方法将偏离交点数从传统方法的11个降至6个,计算时间从0.099μs缩短至0.073μs,效率提升26.0%;在五轴加工应用中,生成G^(2)连续刀具轨迹,干涉检测准确率达99.5%,表面粗糙度控制在0.82μm.
Aiming at the numerical instability of iterative parametric surface intersection in degenerate scenarios,a robustness enhancement method integrating geometric features was proposed.The convex hull aided restart point technique reconstructed oscillatory iteration points by computing substitute points based on the geometric center of control convex hulls to break infinite loops.The oscillation-damping vector decomposition technique suppressed oscillation by decomposing iteration vectors and restraining tangential components.Combined with the parameter-domain affine invariant technique,the regional analysis and Jacobian recalculation mechanism stabilized calculations in tangential contact regions.Experimental results show that on Bézier surface tests,the proposed method reduces deviated intersection points from 11 in traditional methods to 6,shortens computation time from 0.099μs to 0.073μs,and improves efficiency by 26.0%.In five-axis machining applications,it generates G^(2)continuous toolpaths,achieves interference detection accuracy of 99.5%,and controls surface roughness at 0.82μm.
作者
程婷
徐保文
董立春
CHENG Ting;XU Baowen;DONG Lichun(Beijing Center,National Innovation Center for Digital Design and Manufacturing,Beijing 100028,China;Information Technology Center,Aviation Industry Corporation of China,Beijing 100028,China;Avic Digital Co.Ltd.,Beijing 100028,China;School of Mechanical and Aerospace Engineering,Jilin University,Changchun 130021,China)
出处
《华中科技大学学报(自然科学版)》
北大核心
2025年第12期185-192,共8页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家重点研发计划资助项目(2020YFB1709001).
关键词
参数曲面求交
退化
鲁棒性
凸包辅助重启点
振荡抑制矢量分解
仿射不变量
parametric surface intersection
degeneracy
robustness
convex hull aided restart point
oscillation-damping vector decomposition
affine invariant
