A temperature prediction method for optical elements based on Transformer model
-
摘要: 采用Transformer模型来解决多物理场耦合作用下的光学元件实时温度预测难题。试验结果表明，与经验模型法相比，Transformer模型法在均方根误差和平均绝对误差2个指标上分别提升70%和32%；与LSTM法相比，Transformer模型法在均方根误差和平均绝对误差2个指标上分别提升66%和23%；Transformer模型法的决定系数值更加接近1，表明模型的预测结果与真实值吻合度更高。
-
关键词:
- 光学元件 /
- Transformer模型 /
- 实时温度 /
- 温度预测
Abstract: This paper employs a Transformer model to address the challenge of real-time temperature prediction for optical elements under multi-physical field coupling. Experimental results demonstrate that compared to empirical model methods, the Transformer model achieves reductions of 70% and 32% in root mean square error (RMSE) and mean absolute error (MAE), respectively. When compared to LSTM-based methods, the Transformer model reduces RMSE and MAE by 66% and 23%, respectively. Additionally, the coefficient of determination (R2) of the Transformer model approaches 1 more closely, indicating higher consistency between predicted and actual values.-
Key words:
- optical elements /
- Transformer model /
- real-time temperature /
- temperature prediction
-
图 1 基于Transformer的光学元件温度预测模型
Figure 1. Temperature Prediction model For Optical Elements based on Transformer
图 2 多头自注意力机制结构示意图
Figure 2. Architecture Schematic of Multi-Head Self-Attention Mechanism
图 3 三种温度预测方法预测效果对比图
Figure 3. Comparison of Prediction Performance for Three Temperature Prediction Methods
表 1 三种方法的指标对比结果表
Table 1. The results of indicators of three methods
Method δRMSE/℃ δMAE/℃ ${{\text{δ}} }_{{\mathrm{R}}^{2}} $ Tp/s Empirical 2.93 2.23 0.827 0.5 LSTM 1.25 0.97 0.913 1.2 Transformer 0.85 0.75 0.958 0.6 
下载: 导出CSV
-
[1] 邓燕, 许乔, 柴立群, 等. 光学元件亚表面缺陷的全内反射显微检测[J]. 强激光与粒子束, 2009, 21(6):835-840Deng Yan, Xu Qiao, Chai Liqun, et al. Total internal reflection microscopy: a subsurface defects identification technique in optically transparent components[J]. High Power Laser and Particle Beams, 2009, 21(6): 835-840 [2] 任冰强, 黄惠杰, 张维新, 等. 光学元件损伤在线检测装置及实验研究[J]. 强激光与粒子束, 2004, 16(4):465-468Ren Bingqiang, Huang Huijie, Zhang Weixin, et al. Online inspection apparatus and experiments on optics damage[J]. High Power Laser and Particle Beams, 2004, 16(4): 465-468 [3] 赵元安, 邵建达, 刘晓凤, 等. 光学元件的激光损伤问题[J]. 强激光与粒子束, 2022, 34:011004 doi: 10.11884/HPLPB202234.210331Zhao Yuanan, Shao Jianda, Liu Xiaofeng, et al. Tracking and understanding laser damage events in optics[J]. High Power Laser and Particle Beams, 2022, 34: 011004 doi: 10.11884/HPLPB202234.210331 [4] 胡争争, 马留洋, 胡豪. 基于红外和可见光视频的光学元件故障诊断方法[J]. 强激光与粒子束, 2023, 35:089002 doi: 10.11884/HPLPB202335.230040Hu Zhengzheng, Ma Liuyang, Hu Hao. A fault diagnosis method for optical elements based on infrared and visible light videos[J]. High Power Laser and Particle Beams, 2023, 35: 089002 doi: 10.11884/HPLPB202335.230040 [5] Gers F A, Schmidhuber J, Cummins F. Learning to forget: continual prediction with LSTM[J]. Neural Computation, 2000, 12(10): 2451-2471. doi: 10.1162/089976600300015015 [6] Pascanu R, Mikolov T, Bengio Y. On the difficulty of training recurrent neural networks[C]//Proceedings of the 30th International Conference on Machine Learning. 2013: III-1310-III-1318. [7] Cho K, van Merrienboer B, Gulcehre C, et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation[C]//Proceedings of 2014 Conference on Empirical Methods in Natural Language Processing. 2014: 1724-1734. [8] 孟祥福, 石皓源. 基于Transformer模型的时序数据预测方法综述[J]. 计算机科学与探索, 2025, 19(1):45-64 doi: 10.3778/j.issn.1673-9418.2403070Meng Xiangfu, Shi Haoyuan. Survey of Transformer-based model for time series forecasting[J]. Journal of Frontiers of Computer Science & Technology, 2025, 19(1): 45-64 doi: 10.3778/j.issn.1673-9418.2403070 [9] 梁宏涛, 刘硕, 杜军威, 等. 深度学习应用于时序预测研究综述[J]. 计算机科学与探索, 2023, 17(6):1285-1300 doi: 10.3778/j.issn.1673-9418.2211108Liang Hongtao, Liu Shuo, Du Junwei, et al. Review of deep learning applied to time series prediction[J]. Journal of Frontiers of Computer Science & Technology, 2023, 17(6): 1285-1300 doi: 10.3778/j.issn.1673-9418.2211108 [10] 袁烨, 张永, 丁汉. 工业人工智能的关键技术及其在预测性维护中的应用现状[J]. 自动化学报, 2020, 46(10):2013-2030Yuan Ye, Zhang Yong, Ding Han. Research on key technology of industrial artificial intelligence and its application in predictive maintenance[J]. Acta Automatica Sinica, 2020, 46(10): 2013-2030 [11] 冯泽域, 佘少波, 李春煜, 等. 非均匀温度空间中光束内温度预测方法及优化[J/OL]. 真空与低温, 2025. (2025-06-30). http://kns.cnki.net/kcms/detail/62.1125.O4.20250213.1828.002.html.Feng Zeyu, She Shaobo, Li Chunyu, et al. Beam temperature prediction method and optimization in non-uniform temperature space[J/OL]. Vacuum and Cryogenics, 2025. (2025-06-30). http://kns.cnki.net/kcms/detail/62.1125.O4.20250213.1828.002.html. [12] 陈思羽, 徐爱迪, 王贞旭, 等. 基于LSTM算法的玉米籽粒储藏温度预测[J]. 实验技术与管理, 2024, 41(1):57-62Chen Siyu, Xu Aidi, Wang Zhenxu, et al. Corn kernel storage temperature prediction based on the LSTM algorithm[J]. Experimental Technology and Management, 2024, 41(1): 57-62 [13] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. 2017: 6000-6010. [14] Wen Qingsong, Zhou Tian, Zhang Chaoli, et al. Transformers in time series: a survey[C]//Proceedings of the 32nd International Joint Conference on Artificial Intelligence. 2023: 759. [15] 周哲韬, 刘路, 宋晓, 等. 基于Transformer模型的滚动轴承剩余使用寿命预测方法[J]. 北京航空航天大学学报, 2023, 49(2):430-443Zhou Zhetao, Liu Lu, Song Xiao, et al. Remaining useful life prediction method of rolling bearing based on Transformer model[J]. Journal of Beijing University of Aeronautics and Astronautics, 2023, 49(2): 430-443 [16] Chorowski J, Bahdanau D, Serdyuk D, et al. Attention-based models for speech recognition[C]//Proceedings of the 29th International Conference on Neural Information Processing Systems. 2015: 577-585. [17] Bahdanau D, Cho K, Bengio Y. Neural machine translation by jointly learning to align and translate[C]//Proceedings of the 3rd International Conference on Learning Representations. 2015. -
点击查看大图
图(3) / 表(1)
计量
- 文章访问数: 10
- HTML全文浏览量: 7
- PDF下载量: 1
- 被引次数: 0
