首页|基于灵敏度分析的光学花纹设计和加工参数优化

基于灵敏度分析的光学花纹设计和加工参数优化

扫码查看
为同时满足光学花纹工程要求和法规要求,快速找到最优的花纹设计和加工参数,解决从设计到制造需要大量仿真和实验才能进行多目标优化的问题,提出了一种基于灵敏度分析的全过程参数优化方法.在设计阶段,光学花纹设计参数表面粗糙度 0.1~2 mm,光学利用率性能最好;在制造阶段,以HP4A材料为例,以该表面粗糙度范围进行拉丁超立方采样设计,以其最小值构建目标函数,考察切削参数主轴转速n、进给速度Vf和轴向切深ap对表面粗糙度Ra的影响,得到最佳切削参数组合.通过灵敏度分析,并发现Ra与n、Vf呈负相关的关系,与ap呈正相关的关系规律.通过灵敏度分析光学花纹全参数快速优化方法,对比传统实验和仿真方法减少了设计变量个数,提高求解效率 30%,证明该方法有效性.
Optical Pattern Design and Machining Parameter Optimization Based on Sensitivity Analysis
To meet both the requirements of optical pattern engineering and regulatory,and to quickly find the optimal pattern design and processing parameters,thereby resolving the need for extensive simulations and experiments to achieve multi-objective optimization from design to manufacturing,a comprehensive parameter optimization method based on sensitivity analysis is proposed in this paper.The results indicate that during the design phase,the optical pattern achieves optimal optical utilization performance when the surface roughness of the design parameters is maintained between 0.1 mm and 2 mm.During the manufacturing phase,using HP4A material as an example,the design is sampled using Latin hypercube sampling within the specified range of surface roughness,and the objective function is constructed based on its minimum value.The study examined the effects of cutting parameters:spindle speed(n),feed rate(Vf),and axial cutting depth(ap)on surface roughness(Ra),resulting in the identification of the optimal combination of cutting parameters.Through sensitivity analysis,it is shown that surface roughness negatively correlates with spindle speed and feed rate,and positively correlates with axial cutting depth.By using sensitivity analysis to quickly optimize the full parameters of optical patterns,the method reduces the number of design variables and improves solving efficiency by 30%compared to traditional experimental and simulation methods,which proves its effectiveness.

optical patternsmulti-objective optimizationsensitivity analysisLatin hypercube sampling designsurface roughness

应正健、谢正超、李晶

展开 >

华南理工大学机械与汽车工程学院,广东广州 511447

广州卡仕福科技有限公司,广东广州 510700

光学花纹 多目标优化 灵敏度分析 拉丁超立方采样设计 表面粗糙度

2024

光学与光电技术
华中光电技术研究所 武汉光电国家实验室 湖北省光学学会

光学与光电技术

CSTPCD
影响因子:0.351
ISSN:1672-3392
年,卷(期):2024.22(4)