Multi-Physical Intelligent Analysis of the Thermal-Structural-Optical Coupling for Mirror Based on Surrogate Model
The thermal-structural-optical multi-physical coupling analysis (TSO-MCA) of infrared optical systems and their mirrors during the establishment of a cryogenic environment is a complex multi-physical coupling problem involving mechanics, thermal, optics and other factors. Currently, the finite eleme...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10844289/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The thermal-structural-optical multi-physical coupling analysis (TSO-MCA) of infrared optical systems and their mirrors during the establishment of a cryogenic environment is a complex multi-physical coupling problem involving mechanics, thermal, optics and other factors. Currently, the finite element method’s workflow is complex and cumbersome, and it cannot simulate the continuous process of establishing a cryogenic environment. Therefore, a surrogate modeling approach based on a neural network for the TSO-MCA of mirrors has been proposed to analyze the surface shape of optical system mirrors under the action of multi-physical fields. Firstly, a thermal-structural-optical multi-physical integrated chain(TSO-IC) is established by integrating various disciplinary simulation and analysis tools, through which computational data necessary for the surrogate model are acquired. Secondly, a hybrid network combining Transformer and Long Short-Term Memory (LSTM) networks is proposed. The high-precision TSO-MCA surrogate model is obtained through training and K-fold cross validation. Compared with traditional methods, the hybrid network has the smallest symmetric mean absolute percentage error (SMAPE) of 2.674% on the same test set, which indicates that the hybrid network obtains a higher prediction accuracy of the surrogate model. Finally, through the thermal structure-optical coupling analysis of a mirror of an optical system, it is shown that the high precision mirror TSO-MCA based on the surrogate model is 3 orders of magnitude faster than the traditional finite element analysis in terms of computation time, indicating that the surrogate model is more efficient and convenient. |
|---|---|
| ISSN: | 2169-3536 |