Physics-Based Data Generation with SVR Surrogate Modeling for PDMS Thin Film Emissivity Prediction
DOI:
https://doi.org/10.54097/2as25961Keywords:
Radiative Cooling, PDMS Thin Films, Transfer Matrix Method, Support Vector Regression, Surrogate ModelingAbstract
This study combines Transfer Matrix Method (TMM) simulation with Support Vector Regression (SVR) to predict the infrared emissivity of PDMS/SiO₂ thin films across λ=2.0-14.0μm and thicknesses of 100–1000 nm. Physical consistency is enforced during data generation: every TMM spectrum is validated against ε=1-R-T, and points violating |R+T+ε-1|>10-3 are excluded from training. The SVR surrogate, trained on 82,938 validated spectra with an RBF kernel and Halton hyperparameter search, achieves hold-out R2 = 0.959 (RMSE = 0.047, MAE = 0.026) and runs 275× faster than direct TMM evaluation. An overlap-aware multi-region partitioning scheme reduces boundary prediction errors below 1%. The atmospheric-window band (8–13 μm) reaches R2 = 0.985. The complete pipeline, including data, model, and configuration, is publicly archived.
Downloads
References
[1] Mandal, A., Yu, X., Fu, Y., et al. (2018). Scalable and paint-like daytime radiative cooling coatings with high performance. Science, 362(6414), 315–319. https://doi.org/10.1126/science.aau9122
[2] Kou, J., Jurado, Z., Chen, Z., et al. (2020). A dual-layer structure for simultaneous daytime radiative cooling and sunlight harvesting. ACS Applied Materials & Interfaces, 12(27), 30571–30578. https://doi.org/10.1021/acsami.0c07035
[3] Wang, Y., Chen, M., Li, Y., et al. (2021). Pyramid-textured PDMS film for radiative cooling of solar cells. Solar Energy Materials and Solar Cells, 225, 111024. https://doi.org/10.1016/j.solmat.2021.111024
[4] Chen, Y., & Xu, M. (2025). Quantification of spectral emissivity of PDMS thin films based on TMM. Mathematical Modeling and Algorithm Application, 7(2), 49–62.
[5] Raissi, M., Perdikaris, P., & Karniadakis, G. E. (2019). Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 378, 686–707. https://doi.org/10.1016/j.jcp.2018.107078
[6] Yuan, C., Shi, Y., Ba, Z., et al. (2025). Machine learning models for predicting thermal properties of radiative cooling aerogels. Fibers, 11(1), 70. https://doi.org/10.3390/fib11010070
[7] Hu, X., Chen, Y., Wang, Z., et al. (2022). Machine-learning-based Monte Carlo tree search algorithm for Tamm emitter optimization. Optics Express, 30(5), 7890–7902. https://doi.org/10.1364/OE.450226
[8] Chen, J., Wu, Y., & Zhang, H. (2022). Machine learning models for predicting thermal properties of thin films. International Journal of Thermal Sciences, 179, 107654. https://doi.org/10.1016/j.ijthermalsci.2022.107654
[9] Zhang, L., Liu, C., & Wang, H. (2021). Higher-order factorization machine for surrogate modeling of optical thin films. Optics Communications, 503, 126987. https://doi.org/10.1016/j.optcom.2021.126987
[10] Chattopadhyay, U., et al. (2025). Efficient optical coating design using an autoencoder-based neural network model. Journal of Physics: Photonics, 7(1), 015009. https://doi.org/10.1088/2515-7647/ad8381
[11] Vander Wal, M. D., McClarren, R. G., & Humbird, K. D. (2022). Neural network surrogate models for absorptivity and emissivity spectra of multiple elements. Machine Learning: Science and Technology, 3(1), 015009. https://doi.org/10.1088/2632-2153/ac4a82
[12] Park, J., Kim, S., Lee, H., et al. (2022). Surrogate-based optimization of optical thin-film structures using deep learning. Optics Letters, 47(15), 3857–3860. https://doi.org/10.1364/OL.464912
[13] Zhao, R., Zhang, H., Wang, L., et al. (2024). Temperature-dependent infrared optical constants of PDMS films for radiative cooling design. Applied Optics, 63(8), 2156–2164. https://doi.org/10.1364/AO.519647
[14] Ma, T., Wang, H., & Guo, L. J. (2024). OptoGPT: a foundation model for inverse design in optical multilayer thin film structures. Opto-Electronic Advances, 7(7), 240006. https://doi.org/10.29026/oea.20240006
[15] Luce, A., Mahdavi, A., Marquardt, F., et al. (2022). TMM-Fast, a transfer matrix computation package for multilayer thin-film optimization: tutorial. Journal of the Optical Society of America A, 39(6), 1007–1013. https://doi.org/10.1364/JOSAA.454608
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Academic Journal of Applied Sciences
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
