EUVIMEDEUVIMED entdecken
Crossref

P‐4.21: Neural Network‐Based Opto‐Mechanical Analysis of Pancake Optics Under Wide Temperature Ranges

Zheng Wang, Wei Guo, Zi Wang, Guoqiang Lv, Qibin Feng

SID Symposium Digest of Technical Papers · 2026-08 · 57 (S1) · 1260-1263

DOI: 10.1002/sdtp.70003_336 ↗

Kurz gefasst

Fachartikel in der Zeitschrift SID Symposium Digest of Technical Papers, online erschienen am 11.09.2026. Verfasst von Zheng Wang und 4 weiteren Autor:innen. Die Arbeit nennt 4 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Pancake folded optics, the mainstream Virtual Reality/Augmented Reality Head‐Mounted Display optical solution for its compactness and high efficiency, is prone to thermally induced surface form errors and imaging degradation in wide‐temperature environments due to its multi‐fold path, miniaturized components and thermally sensitive packaging. In integrated opto‐mechanical‐thermal analysis, finite element nodal displacement is commonly converted to Zernike coefficients for opto‐mechanical interfacing, yet conventional least squares (LS) method suffers from ill‐conditioning, noise sensitivity and soaring computational cost for high‐order fitting of Pancake even aspheric surfaces with large‐scale nodal data, limiting solving stability and efficiency. We propose an attention‐augmented 1D convolutional neural network to solve Zernike coefficients of thermally induced deformation, with a thermo‐structural finite element model built to acquire nodal sag variation at 20℃, 30℃ and 40℃ as input for coefficient regression. Experimental results show the proposed method reaches a coefficient of determination (R²) of 0.999 on the test set with smaller fitting residuals than the LS baseline; optical simulation confirms high‐temperature defocus term growth is the dominant cause of imaging degradation, providing an efficient closed‐loop evaluation approach for thermal management and structural compensation optimization of Pancake optical modules.

Zitieren

APA

Wang, Z., Guo, W., Wang, Z., Lv, G., & Feng, Q. (2026). P‐4.21: Neural Network‐Based Opto‐Mechanical Analysis of Pancake Optics Under Wide Temperature Ranges. SID Symposium Digest of Technical Papers, 57(S1), 1260-1263. https://doi.org/10.1002/sdtp.70003_336

Vancouver

Wang Z, Guo W, Wang Z, Lv G, Feng Q. P‐4.21: Neural Network‐Based Opto‐Mechanical Analysis of Pancake Optics Under Wide Temperature Ranges. SID Symposium Digest of Technical Papers. 2026;57(S1):1260-1263. doi:10.1002/sdtp.70003_336

BibTeX
@article{Wang2026,
  author = {Wang, Zheng and Guo, Wei and Wang, Zi and Lv, Guoqiang and Feng, Qibin},
  title = {P‐4.21: Neural Network‐Based Opto‐Mechanical Analysis of Pancake Optics Under Wide Temperature Ranges},
  journal = {SID Symposium Digest of Technical Papers},
  year = {2026},
  volume = {57},
  number = {S1},
  pages = {1260--1263},
  doi = {10.1002/sdtp.70003_336}
}
Kostenlos starten →