AI 中文总结
研究旨在解决大质量恒星参数计算的计算瓶颈，通过计算FASTWIND合成光谱训练神经网络，开发SpecFANN包。该方法能快速准确获取恒星参数，相比传统方法计算时间大幅减少，证明神经网络可突破现有计算限制。
AI 中文摘要
大质量恒星非常重要：它们是早期宇宙的有力探测器，在宿主环境的化学和力学演化中起关键作用，其最终产物有助于研究宇宙中最极端的物理。获取大质量恒星样本的准确恒星和表面参数对理解其演化及对周围环境的影响至关重要。随着即将到来的光谱巡天产生大量数据，计算限制可能成为瓶颈。为解决此问题并大幅减少计算时间，我们旨在为FASTWIND辐射传输和光谱合成代码开发强大的模拟器，探索因计算成本此前不可行的替代拟合方法。我们计算了一组OB型恒星的FASTWIND合成光谱，训练神经网络来模拟这些模型，还开发了开源Python包SpecFANN，为用户提供可与这些或其他用户生成的神经网络一起使用的拟合方法套件。大多数训练的神经网络对光球线的平均精度优于约0.01 - 0.1%，对风线优于约0.1 - 1%。SpecFANN能为52颗早型恒星样本获得与文献一致的稳健且准确的恒星参数。与依赖实时FASTWIND计算的替代技术相比，使用SpecFANN能在约1/360,000的时间内实现相同拟合。我们证明神经网络为解决当前热星大气分析和恒星参数确定方法的计算限制提供了可行途径。
英文摘要
The importance of massive stars cannot be overstated: they are powerful probes of the early universe, play a vital role in the chemical and mechanical evolution of their host environments and their end products allow us to study the most extreme physics in the universe. Obtaining accurate stellar and surface parameters for large samples of massive stars is vital to our understanding of how they evolve, and how their births, lives and deaths affect their surroundings. With the large volume of data expected from upcoming spectroscopic surveys, computational limitations will likely be the most important bottleneck impeding our progress. To address this and dramatically decrease computing times, we aim to develop a robust emulator for the FASTWIND radiative transfer and spectral synthesis code. Additionally, we aim to explore alternative fitting methods that have not been feasible until now due to computational costs. We calculate a set of FASTWIND synthetic spectra of OB-type stars, and we train a collection of neural networks to emulate these models. We also develop the open-source python package SpecFANN, which provides users with a suite of fitting methods that can be used with these or other user-generated neural networks. The majority of the trained neural networks reach average accuracies of better than ~0.01-0.1% for photospheric lines and better than ~0.1-1% for wind lines. SpecFANN is able to obtain robust and accurate stellar parameters that are consistent with the literature for a sample of 52 early-type stars. Using SpecFANN we find that we can achieve the same fit in ~1/360,000 of the time when compared to alternative techniques that rely on on-the-fly FASTWIND computations. We have demonstrated that neural networks offer a viable path forward to address the computational limitations of our current atmosphere analysis and stellar parameter determination methods for hot stars.
CommentsAccepted for publication in A&A; 12 pages (+2 appendix pages), 9 figures (+2 appendix figures)
