照明工程学报2025，Vol.36Issue(2)：82-87,6.DOI:10.3969/j.issn.1004-440X.2025.02.011
基于CPO-BP模型的TIR透镜的设计方法
Design Method of TIR Lens Based on CPO-BP Model
于鲁平 1李会明 2李伟康 2冶振2
作者信息
- 1. 中国人民解放军93170 部队,陕西 西安 710000
- 2. 兰州万里航空机电有限责任公司,甘肃 兰州 730050
- 折叠
摘要
Abstract
The structure of the TIR lens has an significant impact on the light intensity and scattering angle of the LED light source passing through it.A BP neural network model optimized by the Crested Porcupine Optimizer CPO is established by taking the structural parameters of the TIR lens as input variables and the light intensity and scattering angle as output variables.The results show that,compared to the BP neutral network,the CPO-optimized BP neural network improved the average absolute percentage error(MAPE)for light intensity and scattering angle prediction by 97.85%and 94.3%,respectively.The CPO-BP model demonstrates good effectiveness,accuracy and scalability.关键词
TIR透镜/冠豪猪优化算法(CPO)/BP神经网络/配光设计Key words
TIR lens/crested porcupine optimizer(CPO)/BP neural network/light distribution design分类
信息技术与安全科学引用本文复制引用
于鲁平,李会明,李伟康,冶振..基于CPO-BP模型的TIR透镜的设计方法[J].照明工程学报,2025,36(2):82-87,6.
