E-mail Alert
RSS
| Citation: |
Zhang Z Y, Chang J, Huang Y F, et al. Reinforcement learning-based stray light suppression study for space-based gravitational wave detection telescope system[J]. Opto-Electron Eng, 2024, 51(2): 230210. doi: 10.12086/oee.2024.230210
|
Reinforcement learning-based stray light suppression study for space-based gravitational wave detection telescope system
-
Abstract
In gravitational wave telescopes, the energy of the collected space target light signals is dwarfed by the energy of stray light, necessitating robust stray light suppression for reliable telescope operation. Due to the inherent unpredictability of scattered light and the intricate nature of opto-mechanical systems, the formulation of stray light suppression strategies often involves complex mathematical modeling, substantial expertise, and iterative simulations. This paper introduces a Reinforcement Learning-based approach to devise the stray light suppression scheme within a Monte Carlo ray tracing environment, specifically for space gravitational wave telescope systems. Our empirical findings confirm the efficacy of this methodology in generating effective stray light suppression strategies, yielding favorable suppression performance. This study contributes a novel, efficient, and adaptable solution to the stray light challenges faced in space gravitational wave detection as well as other high-precision optical systems, thereby holding extensive applicative promise. -
-
References
[1] 王智, 马军, 李静秋. 空间引力波探测计划-LISA系统设计要点[J]. 中国光学, 2015, 8(6): 980−987. doi: 10.3788/CO.20150806.0980 Wang Z, Ma J, Li J Q. Space-based gravitational wave detection mission: design highlights of LISA system[J]. Chin Opt, 2015, 8(6): 980−987. doi: 10.3788/CO.20150806.0980 [2] 王智, 沙巍, 陈哲, 等. 空间引力波探测望远镜初步设计与分析[J]. 中国光学, 2018, 11(1): 131−151. doi: 10.3788/CO.20181101.0131 Wang Z, Sha W, Chen Z, et al. Preliminary design and analysis of telescope for space gravitational wave detection[J]. Chin Opt, 2018, 11(1): 131−151. doi: 10.3788/CO.20181101.0131 [3] 王登峰, 姚鑫, 焦仲科, 等. 面向天基引力波探测的时间延迟干涉技术[J]. 中国光学, 2021, 14(2): 275−288. doi: 10.37188/CO.2020-0098 Wang D F, Yao X, Jiao Z K, et al. Time-delay interferometry for space-based gravitational wave detection[J]. Chin Opt, 2021, 14(2): 275−288. doi: 10.37188/CO.2020-0098 [4] Nardello M, Lintz M. Effective cross sections for stray light calculations in laser interferometry: application to LISA science interferometer[J]. Proc SPIE, 2021, 11852: 118523Q. doi: 10.1117/12.2599618 [5] 王小勇,白绍竣,张倩,等. 空间引力波探测望远镜研究进展[J]. 光电工程, 2023, 50(11): 230219. doi: 10.12086/oee.2023.230219 Wang X Y, Bai S J, Zhang Q, et al. Research progress of telescopes for space-based gravitational wave missions[J]. Opto-Electron Eng, 2023, 50(11): 230219. doi: 10.12086/oee.2023.230219 [6] Livas J C, Sankar S R. Optical telescope system-level design considerations for a space-based gravitational wave mission[J]. Proc SPIE, 2016, 9904: 99041K. doi: 10.1117/12.2233249 [7] 张耘豪,钟哲强,张彬. 空间引力波探测系统中超光滑光学元件表面散射特性分析[J]. 光电工程, 2023, 50(11): 230222. doi: 10.12086/oee.2023.230222 Zhang Y H, Zhong Z Q, Zhang B. Analysis of surface scattering characteristics of ultra-smooth optical components in gravitational wave detection system[J]. Opto-Electron Eng, 2023, 50(11): 230222. doi: 10.12086/oee.2023.230222 [8] Lapeyrolerie M, Chapman M S, Norman K E A, et al. Deep reinforcement learning for conservation decisions[J]. Methods Ecol Evol, 2022, 13(11): 2649−2662. doi: 10.1111/2041-210X.13954 [9] Heuillet A, Couthouis F, Díaz-Rodríguez N. Explainability in deep reinforcement learning[J]. Knowl-Based Syst, 2021, 214: 106685. doi: 10.1016/j.knosys.2020.106685 [10] Jordan M I, Mitchell T M. Machine learning: trends, perspectives, and prospects[J]. Science, 2015, 349(6245): 255−260. doi: 10.1126/science.aaa8415 [11] Liu H D, Zhou H C, Wang D D, et al. Performance comparison of two Monte Carlo ray-tracing methods for calculating radiative heat transfer[J]. J Quant Spectrosc Radiat Transfer, 2020, 256: 107305. doi: 10.1016/j.jqsrt.2020.107305 [12] Yarahmadi M, Mahan J R, Priestley K J. Uncertainty analysis and experimental design in the Monte Carlo ray-trace environment[J]. J Heat Transfer, 2019, 141(3): 032701. doi: 10.1115/1.4042365 [13] Howell J R. The Monte Carlo method in radiative heat transfer[J]. J Heat Transfer, 1998, 120(3): 547−560. doi: 10.1115/1.2824310 [14] Duan X Y, He C T, Lin X X, et al. Quasi-Monte Carlo ray tracing algorithm for radiative flux distribution simulation[J]. Sol Energy, 2020, 211: 167−182. doi: 10.1016/j.solener.2020.09.061 -
Overview
In the field of space gravitational wave detection, the functionality of telescopes is critically hampered by the presence of stray light. This stray light, significantly overpowering the energy of collected space-target optical signals, poses a substantial challenge, necessitating robust stray light suppression for the telescope's reliable operation. Traditionally, this suppression has been achieved through complex mathematical modeling, extensive expertise, and iterative simulations, given the inherent unpredictability of scattered light and the intricate nature of optical-mechanical systems.
This study introduces a novel approach to stray light suppression by integrating reinforcement learning, a key area in machine learning. Reinforcement learning optimizes problem-solving strategies through an intelligent agent that learns to maximize cumulative rewards from its environment to achieve specific objectives. By applying this methodology, we address the limitations of traditional stray light suppression scheme formulation.
Our approach employs reinforcement learning to develop a strategy for stray light suppression within a Monte Carlo ray tracing-based environment, tailored for Space Gravitational Wave Detection Telescope systems. This method is particularly adept at identifying and mitigating stray light paths caused by surface scattering, a prevalent issue in optical systems.
A comparative analysis was conducted using the specialized stray light analysis software to evaluate the effectiveness of traditional methods against those developed through reinforcement learning. The results reveal that reinforcement learning-based schemes surpass traditional methods in identifying and mitigating stray light paths, especially those caused by surface scattering. Compared with the traditional stray light analysis and suppression, it reduces the designer's manual iteration process based on experience and ray tracing results due to the high standard of stray light suppression brought by the space gravitational wave detection telescope system. The scattering caused by surface properties is considered, and the stray light suppression effect is improved.
The empirical findings from this study confirm the superiority of reinforcement learning in formulating effective stray light suppression measures for space gravitational wave detection telescope systems. The approach not only achieves superior suppression outcomes but also introduces an efficient, flexible, and innovative solution to the challenges of stray light in space gravitational wave detection and other high-precision optical systems.
-
Access History
Figures(13)
Tables(1)
Article Metrics
Export File
Citation
Zhang Z Y, Chang J, Huang Y F, et al. Reinforcement learning-based stray light suppression study for space-based gravitational wave detection telescope system[J]. Opto-Electron Eng, 2024, 51(2): 230210. doi: 10.12086/oee.2024.230210
Format
Content
DownLoad:
-
Figure 1.
The optical path diagram of space gravitational wave detection telescope system
-
Figure 2.
The process of traditional stray light analysis and suppression
-
Figure 3.
The length diagram of the baffle at different angles
-
Figure 4.
The optomechanical structure under the traditional design method
-
Figure 5.
RL network model. (a) Env, Actor, Critic; (b) Ray tracing model for environment
-
Figure 6.
The reward fuction
-
Figure 7.
The initial stray light suppression effect of the model
-
Figure 8.
The stray light suppression effect of the traditional stray light suppression scheme
-
Figure 9.
The stray light path under the traditional stray light suppression scheme
-
Figure 10.
The RL running results
-
Figure 11.
The optical-mechanical structure of RL stray light suppression scheme
-
Figure 12.
Comparison of the stray light suppression effects of different stray light suppression schemes
- Figure .
