面向AI4S的天文数智化服务体系构建与实践——以国家天文科学数据中心为例

doi:10.11871/jfdc.issn.2096-742X.2026.04.001

数据与计算发展前沿 ›› 2026, Vol. 8 ›› Issue (4): 1-18.

CSTR: 32002.14.jfdc.CN10-1649/TP.2026.04.001

doi: 10.11871/jfdc.issn.2096-742X.2026.04.001

面向AI4S的天文数智化服务体系构建与实践——以国家天文科学数据中心为例

崔辰州^1,^2,^*(),许允飞^1,²,樊东卫^1,²,李长华^1,²,李珊珊^1,²,米琳莹^1,²,甘庆波^1,²,赵公博^1,²,China-VO团队

¹ 中国科学院国家天文台，北京 100101
² 国家天文科学数据中心，北京 100101

收稿日期:2026-02-05 出版日期:2026-08-20 发布日期:2026-08-21
通讯作者: *崔辰州（E-mail：ccz@nao.cas.cn）
作者简介:崔辰州，中国科学院国家天文台，研究员，国家天文科学数据中心，常务副主任，主要研究领域为天文信息学、虚拟天文台、人工智能驱动的科学研究。
本文承担的工作为整体研究框架及论文撰写。
CUI Chenzhou is a professor at the National Astronomical Observatories, Chinese Academy of Sciences, and executive deputy director of the National Astronomical Data Center. His main research areas include astroinformatics, virtual observatory, and AI for Science.
In this paper, he is mainly responsible for overall research framework and manuscript writing.
E-mail: ccz@nao.cas.cn
基金资助:
国家重点研发计划(2022YFF0711500);国家自然科学基金(12573111);国家自然科学基金(12403102);国家自然科学基金(12373110);国家自然科学基金(12273077);国家自然科学基金(12103070);中国科学院战略性先导科技专项(XDB0550101)

Building an AI4S-Oriented Intelligent Data Service System for Astronomy: Practices from the National Astronomical Data Center

CUI Chenzhou^1,^2,^*(),XU Yunfei^1,²,FAN Dongwei^1,²,LI Changhua^1,²,LI Shanshan^1,²,MI Linying^1,²,GAN Qingbo^1,²,ZHAO Gongbo^1,²,China-VO Collaboration

¹ National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, China
² National Astronomical Data Center, Beijing 100101, China

Received:2026-02-05 Online:2026-08-20 Published:2026-08-21

摘要/Abstract

摘要：

【目的】 天文学已全面步入数据密集型科学发现时代，而人工智能驱动的科学研究（AI4S）范式的兴起，对作为国家战略科技力量组成部分的科学数据中心提出了前所未有的能力提升需求。国家天文科学数据中心（NADC）面对海量异构数据的治理压力、传统数据服务模式的效率瓶颈以及支撑前沿AI4S创新的使命，系统性地规划构建了以“数智化”为核心特征的下一代天文数据服务体系。【方法】 首先完整阐述该体系“三位一体”的总体设计：以应用程序接口（API）和模型上下文协议（MCP）为双核构建标准化与智能化兼备的“数据服务层”；以深度集成数据、软件、算力、模型资源的云端科研平台为承载，打造一站式“科研支撑层”；通过深度嵌入重大科学工程、革新科教融合模式、开展学科和人才建设、拓展公众科学及深化国际合作，培育开放协同的“应用生态层”。【结果】 剖析该体系在LAMOST天体光谱数据库、AI智能体驱动的天关卫星时域天文发现及公众科学项目中的成功实践，展示其在提升数据智能化供给、降低前沿技术使用门槛、赋能天文学发现等方面的显著成效。【结论】 NADC的探索为大数据与人工智能时代国家科学数据中心的功能重塑、能力建设与服务模式创新提供了具有示范意义的系统性解决方案。

关键词: 国家天文科学数据中心（NADC）, 数据密集型科学, 人工智能驱动的科学研究（AI4S）, 虚拟天文台（VO）, 数智化服务, 开放科学, 模型上下文协议（MCP）, 智能体

Abstract:

[Objective] Astronomy has fully entered the era of data-intensive scientific discovery. Meanwhile, the rapid emergence of the artificial intelligence-driven scientific research paradigm, namely AI for Science (AI4S), has imposed unprecedented requirements for enhancing the capability of scientific data centers as integral components of national strategic scientific and technological forces. Confronted with the challenges of governing massive and heterogeneous datasets, the efficiency bottlenecks of traditional data service models, and the mission of supporting cutting-edge AI4S innovation, the National Astronomical Data Center (NADC) has systematically planned and implemented a next-generation astronomical data service system characterized by “Digital-Intelligent Integration”. [Methods] This paper first presents a comprehensive description of the system's “trinity” overall architecture. A standardized and intelligent data service layer is established with application programming interfaces (APIs) and the Model Context Protocol (MCP) as dual cores. A one-stop research support layer is built upon the cloud-based scientific research platform, the China-VO Science Platform, which deeply integrates data, software, computing resources, and models. An open and collaborative application ecosystem layer is cultivated through deep engagement with major scientific projects, innovation in science-education integration models, disciplinary and talent development, expansion of citizen science initiatives, and strengthened international collaboration. [Results] The paper focuses on in-depth analyses of successful practices within this system, including standardized and intelligent services for the LAMOST astronomical spectral database, AI agent-driven time-domain astronomical discovery with EP satellites, and citizen science programs. These cases demonstrate the system's significant effectiveness in enhancing intelligent data provision, lowering the barriers to adopting advanced AI technologies, and empowering discoveries in time-domain astronomy. [Conclusions] The explorations of NADC provide a systematic and demonstrative solution for functional transformation, capacity building, and service model innovation of national scientific data centers in the era of big data and artificial intelligence.

Key words: National Astronomical Data Center (NADC), data-intensive scientific discovery, AI for Science (AI4S), Virtual Observatory (VO), intelligent data service, open science, Model Context Protocol (MCP), agent

崔辰州, 许允飞, 樊东卫, 李长华, 李珊珊, 米琳莹, 甘庆波, 赵公博, China-VO团队. 面向AI4S的天文数智化服务体系构建与实践——以国家天文科学数据中心为例[J]. 数据与计算发展前沿, 2026, 8(4): 1-18.

CUI Chenzhou, XU Yunfei, FAN Dongwei, LI Changhua, LI Shanshan, MI Linying, GAN Qingbo, ZHAO Gongbo, China-VO Collaboration. Building an AI4S-Oriented Intelligent Data Service System for Astronomy: Practices from the National Astronomical Data Center[J]. Frontiers of Data and Computing, 2026, 8(4): 1-18, https://cstr.cn/32002.14.jfdc.CN10-1649/TP.2026.04.001.

图/表 8

图1

图2

表1

表2

图3

图4

图5

图6

参考文献 50

[1]	CUI X Q, ZHAO Y H, CHU Y Q, et al. The Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST)[J]. Research in Astronomy and Astrophysics, 2012, 12:1197. https://doi.org/10.1088/1674-4527/12/9/003. doi: 10.1088/1674-4527/12/9/003
[2]	NAN R D, LI D, JIN C J, et al. The Five-Hundred-Meter Aperture Spherical Radio Telescope (FAST) Project[J]. Int J Mod Phys D, 2011, 20:989. https://arxiv.org/abs/1105.3794. doi: 10.1142/S0218271811019335
[3]	IVEZI´C ˇZ, KAHN S M, TYSON J A, et al. LSST: From Science Drivers to Reference Design and Anticipated Data Products[J]. ApJ, 2019, 873: 111. https://doi.org/10.3847/1538-4357/ab042c.
[4]	HALL P J, SCHILIZZI R T, DDWDNEY P E F, et al. The square kilometer array (SKA) radio telescope: Progress and technical directions[C]. In: URSI Radio Science Bulletin, Sep. 2008, 326: 4-19. https://doi.org/10.23919/URSIRSB.2008.7909649.
[5]	YUAN W, ZHANG C, FENG H, et al. Einstein Probe - a small mission to monitor and explore the dynamic X-ray Universe[J]. arXiv e-prints, 2015. https://doi.org/10.48550/arXiv.1506.07735.
[6]	CSST Collaboration, GONG Y, MIAO H, et al. Introduction to the Chinese Space Station Survey Telescope (CSST)[J]. SCIENCE CHINA Physics, Mechanics & Astronomy, 2026, 69(3): 239501. https://doi.org/10.1007/s11433-025-2809-0.
[7]	Euclid Collaboration, MELLIER Y, ABDURRO’UF, et al. Euclid. I. Overview of the Euclid mission[J]. arXiv:2405.13491, 2024. https://doi.org/10.1051/00-04-6361/202450810.
[8]	GREEN J, SCHECHTER P, BALTAY C, et al. Wide-Field InfraRed Survey Telescope (WFIRST) Final Report[J/OL]. 2012. https://arxiv.org/abs/1208.4012.
[9]	DOE Office of Science. THE AMERICAN SCIENCE CLOUD (AmSC). DOE National Laboratory Program Announcement Number: LAB 25-3555. 2025[EB/OL]. https://science.osti.gov/-/media/grants/pdf/lab-announcements/2025/LAB-25-3555.pdf.
[10]	赵永恒, 崔辰州. 天文数据处理与虚拟天文台[M]. 北京: 中国科学技术出版社, 2024: 275-432. ISBN: 978-7-5236-0827-2.
[11]	QUINN P J, BARNES D G, CSABAI I, et al. The International Virtual Observatory Alliance: recent technical developments and the road ahead[C]. In: P. J. Quinn & A. Bridger (eds.), Proceedings of the SPIE, Optimizing Scientific Return for Astronomy through Information Technologies, 2004, 5493: 137. https://doi.org/10.1117/12.551247.
[12]	GINSBURG A, SIPOCZ B M, BRASSEUR C E, et al. astroquery: An Astronomical Web-querying Package in Python[J]. The Astronomical Journal, 2019, Volume 157, Issue 3, article id. 98, 7 pp. https://doi.org/10.3847/1538-3881/aafc33.
[13]	Astropy Collaboration, ROBITAILLE T P, TOLLERUD E J, et al. Astropy: A community Python package for astronomy[J]. Astronomy & Astrophysics, 2013, Volume 558, id.A33, 9 pp. https://doi.org/10.1051/0004-6361/201322068.
[14]	GENOVA F, EGRET D, BIENAYME O, et al. The CDS information hub. On-line services and links at the Centre de Données astronomiques de Strasbourg[J]. Astronomy and Astrophysics Supplement, 2000, 143: 1-7. https://doi.org/10.1051/aas:2000333. doi: 10.1051/aas:2000333
[15]	CUI C, TAO Y, LI C, et al. Towards an astronomical science platform: Experiences and lessons learned from Chinese Virtual Observatory[J]. Astronomy and Computing, 2020, 32:100392. https://doi.org/10.1016/j.ascom.2020.100392. doi: 10.1016/j.ascom.2020.100392
[16]	ARVISET C, NAVARRO V, BASSO D, et al. ESA DATALABS: Towards a Collaborative E-Science Platform for ESA[C]. In:5th Planetary Data Workshop & Planetary Science Informatics & Analytics, held virtually June 28-July 2, 2021. LPI Contribution No. 2549, id. 701. June 2021. https://ui.adsabs.harvard.edu/abs/2021LPICo2549.7014A.
[17]	GAUDET S, DOWLER P, GOLIATH S, et al. The Canadian Advanced Network For Astronomical Research[C]. In: Astronomical Data Analysis Software and Systems XVIII ASP Conference Series, 2009, Vol. 411, proceedings of the conference held 2-5 November 2008 in Québec City, QC, Canada. Edited by David A. Bohlender, Daniel Durand, and Patrick Dowler. San Francisco: Astronomical Society of the Pacific, 2009: 185. https://ui.adsabs.harvard.edu/abs/2009ASPC.411.185G.
[18]	TAGHIZADEH-POPP M, KIM J W, LEMSON G, et al. SciServer: A science platform for astronomy and beyond[J]. Astronomy and Computing, 2020, Volume 33, article id. 100412. https://doi.org/10.1016/j.ascom.2020.100412.
[19]	O’MULLANE W, ECONOMOU F, HUANG F, et al. Rubin Science Platform on Google: the story so far[C]. In: ADASS XXXI, ASP Conference Series, 2024, Vol. 535, Proceedings of the conference held 24-28 October 2021 at Cape Town, South Africa. San Francisco: Astronomical Society of the Pacific:227. https://arxiv.org/pdf/2111.15030.
[20]	Gaia Collaboration, VALLENARI A, BROWN A G A, et al. Gaia Data Release 3. Summary of the content and survey properties[J]. Astronomy & Astrophysics, 2023, Volume 674, id.A1, 22 pp. https://doi.org/10.1051/0004-6361/202243940.
[21]	ABAZAJIAN K, ADELMAN-MCCARTHY J K, Agüeros M A, et al. The First Data Release of the Sloan Digital Sky Survey[J]. ApJS, 2003, 126(4): 2081. https://doi.org/10.1086/378165.
[22]	ZHANG H L, WWANG J, ZHANG Y Z, et al. Review of artificial intelligence applications in astronomical data processing[J]. Astronomical Techniques and Instruments, 2024, 1(1): 1-15. https://doi.org/10.61977/ati2024001.
[23]	李国杰. 智能化科研(AI4R):第五科研范式[J]. 中国科学院院刊, 2024, 39(1): 1-9. https://doi.org/10.16418/j.issn.1000-3045.20231007002.
[24]	鄂维南. AI助力打造科学研究新范式[J]. 中国科学院院刊, 2024, 39(1): 10-16. https://doi.org/10.16418/j.issn.1000-3045.20231224001.
[25]	臧冀原, 季桓永, 黄庆学. 数智化赋能传统产业转型升级[J]. 中国科学院院刊, 2024, 39 (7):1183-1190. https://doi.org/10.16418/j.issn.1000-3045.202404070-04.
[26]	ZHANG Z D, ZHANG Y Y, YAO H D, et al. Xiwu: A Basis Flexible and Learnable LLM for High Energy Physics[J]. arXiv e-print, 2024. https://arxiv.org/abs/2404.08001.
[27]	米琳莹, 崔辰州, 樊东卫, 等. 国家天文科学数据中心发展思路浅析[J]. 农业大数据学报, 2019, 1(4): 37-45. https://doi.org/10.19788/j.issn.2096-6369.190-404. doi: 10.19788/j.issn.2096-6369.190404
[28]	WILKINSON M, DUMONTIER M, AALBERSBERT I, et al. The FAIR Guiding Principles for scientific data management and stewardship[J]. Sci Data 3, 2016, 160018. https://doi.org/10.1038/sdata.2016.18.
[29]	ROY T F. Architectural Styles and the Design of Network-based Software Architectures, CHAPTER 5: Representational State Transfer (REST)[D]. PhD DISSERTATION of UNIVERSITY OF CALIFORNIA, IRVINE, 2000. https://roy.gbiv.com/pubs/dissertation/top.htm.
[30]	ANTHROPIC. Introducing the Model Context Protocol[EB/OL]. 2024. https://www.anthropic.com/news/model-context-protocol.
[31]	UNESCO. UNESCO Recommendation on Open Science[EB/OL]. 2021. https://doi.org/10.54677/MNMH8546.
[32]	LI S S, CUI C Z, FAN D W, et al. Astronomy Citizen Science Projects in China. In: Integrating Science Education, Scientific Discovery, and AI Empowerment[C]. EDULEARN25 Proceedings, 2025: 1688-1697. https://doi.org/10.21125/edulearn.2025.0512.
[33]	ZUO X X, TAO Y H, HUANG Y, et al. FALCO: Foundation Model of Astronomical Light Curves for Time Domain Astronomy. Implementation and Applications on Kepler Data[J]. The Astronomical Journal, 2026, Volume 171, Issue 1, id.10, 19 pp. https://doi.org/10.3847/1538-3881/ae1467.
[34]	ZHAO X S, HUANG Y, XUE G, et al. SpecCLIP: Aligning and Translating Spectroscopic Measurements for Stars[J]. The Astrophysical Journal, 2026, 98:189. doi.org/10.3847/1538-4357/ae2c7e.
[35]	LI S S, LI C Q, LI C H, et al. Mining Double-line Spectroscopic Candidates in the LAMOST Medium-resolution Spectroscopic Survey Using a Human-AI Hybrid Method[J]. ApJS, 2025, 276,11: 14. https://doi.org/10.3847/1538-4365/ad9010.
[36]	YANG X Y, LI C H, ZHANG Y X, et al. Quasar candidate selection and redshift estimation using J-PLUS DR3 data[J]. MNRAS, 2025, 542:2910-2924. https://doi.org/10.1093/mnras/staf1427. doi: 10.1093/mnras/staf1427
[37]	WEI S R, LI C H, ZHANG Y X, et al. Photometric redshift estimation for emission line galaxies of DESI Legacy Imaging Surveys by CNN-MLP[J]. PASA, 2025, 42: e092. https://doi.org/10.1017/pasa.2025.10-056.
[38]	MA X Y, ZHANG Y X, ZHANG J Y, et al. Search for Young Stellar Objects within 4XMM-DR13 Using CatBoost and SPE[J]. AJ, 2024, 168: 210. https://doi.org/10.3847/1538-3881/ad781c. doi: 10.3847/1538-3881/ad781c
[39]	陈学雷, 阎敬业, 徐怡冬, 等. 宇宙黑暗时代探路者——鸿蒙计划[J]. 空间科学学报, 2023, 43(1): 43-59. https://doi.org/10.11728/cjss2023.01.220104001.
[40]	邓元勇, 周桂萍, 代树武, 等. 太阳极轨天文台[J]. 科学通报, 2023, 68 (4): 298-308. https://doi.org/10.1360/TB-2022-0674.
[41]	LI G C, HU M K, LI W X, et al. A Shock Flash Breaking Out of a Dusty Red Supergiant[J]. Nature, 2023, 627(8005):754-758. https://doi.org/10.1038/s4-1586-023-06843-6. doi: 10.1038/s41586-023-06843-6
[42]	FABRIZIO G, YAGO A, LUCA C, et al. Hierarchical Progressive Survey (HiPS) format: Moving from visualization to scientific analysis[J]. A&A, 2025, Volume 703, id.A194, 10 pp. https://doi.org/10.1051/0004-6361/202555224.
[43]	RUTLEDGE R E. The Astronomer's Telegram: A Web-based Short-Notice Publication System for the Professional Astronomical Community[J]. PASP, 1998, 110(748): 754-756. https://doi.org/10.1086/31-6184. doi: 10.1086/pasp.1998.110.issue-748
[44]	DEY A, SCHLEGEL D J, LANG D, et al. Overview of the DESI Legacy Imaging Surveys[J]. The Astronomical Journal, 2019, Volume 157, Issue 5, article id. 168, 29 pp. https://doi.org/10.3847/1538-3881/ab-089d.
[45]	SCHULMAN J, WOLSKI F, DHARIWAL P, et al. Proximal Policy Optimization Algorithms[J/OL]. 2017. https://arxiv.org/abs/1707.06347.
[46]	崔辰州. 公众科学, 让每个人享受探索宇宙的乐趣[J]. 中国国家天文, 2025, 19(11): 16-17.
[47]	LI S S, FAN D W, GAO X, et al. Popular Supernova Project: A Citizen Science Program Based on Amateur Astronomical Observations[C]. In INTED2016 Proceedings, 2016: 7561-7569. DOI:10.21125/inted.2016.0780.
[48]	李珊珊, 崔辰州. 公众科学, 让兴趣爱好连接真实科学发现[J]. 中国国家天文, 2025, 19(11): 34-41.
[49]	高兴. 你我的星辰大海: 中国公众科学之路与PSP项目的十年探索[J]. 中国国家天文, 2025, 19(11): 42-47.
[50]	王艺. 从赋能到重塑,AI Agent 构建企业智能新引擎[EB/OL]. 云栖战略参考, 2025. https://developer.aliyun.com/article/1689470?spm=a2c6h.12873581.te-chnical-group.dArticle1689470.13b42cb4r21y6I.

协议名称	主要用途
VOTable	IVOA标准数据交换格式，用于存储和传输LAMOST星表、光谱元数据与观测参数，支持多平台、多软件的数据互通与解析，保障数据结构的标准化与可重用性。
SAMP	简单应用消息协议，用于天文软件的互操作，支持LAMOST数据在TOPCAT、Aladin等主流天文软件间的实时数据共享与协同分析，实现检索结果的跨软件同步可视化。
ConeSearch	天球锥形区域检索协议，支持用户通过指定赤经、赤纬与搜索半径，快速检索目标天区内的LAMOST天体数据。
SSAP	简单光谱访问协议，专门用于光谱数据的标准化检索与访问，返回光谱数据及配套元信息。
TAP	数据表访问协议，支持用户通过天文数据查询语言ADQL语句对LAMOST光谱数据库进行复杂检索，可实现多条件组合筛选、跨表关联查询，满足深度数据挖掘需求。

API名称	功能描述	访问方式	访问地址（URL）
数据版本查询	获取LAMOST所有已公开的数据版本列表	GET请求	https://www.lamost.org/openapi/dr_versions
表数据查询	针对指定数据表执行结构化条件查询	POST请求	https://www.lamost.org/openapi/{dr_version}/{sub_version}/query/{table_name}
SQL语句查询	支持通过ADQL/SQL语句执行复杂的自定义查询	GET请求	https://www.lamost.org/openapi/{dr_version}/{sub_version}/sql
星表文件下载	下载指定分辨率光谱的全量星表文件	GET请求	https://www.lamost.org/openapi/{dr_version}/{sub_version}/{resolution}/catalog
光谱FITS获取	获取指定天体的光谱数据FITS文件	GET请求	https://www.lamost.org/openapi/{dr_version}/{sub_version}/{resolution}/spectrum/fits
光谱CSV转换	将光谱数据转换为CSV格式	GET请求	https://www.lamost.org/openapi/{dr_version}/{sub_version}/{resolution}/spectrum/fits2csv

面向AI4S的天文数智化服务体系构建与实践——以国家天文科学数据中心为例

Building an AI4S-Oriented Intelligent Data Service System for Astronomy: Practices from the National Astronomical Data Center

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 50

相关文章 10

编辑推荐

Metrics

本文评价

[1]	耿桦, 王莉娜, 高满达, 王文彬, 李亚巍, 游博. 基于多方法融合的煤炭供应链建模仿真研究综述：方法、应用及发展趋势[J]. 数据与计算发展前沿, 2026, 8(4): 203-213.
[2]	周晓芳, 鲍彦荣, 顾荣杰, 王昕. 基于多智能体协作和复杂网络理论的贸易事件知识图谱构建[J]. 数据与计算发展前沿, 2026, 8(4): 234-245.
[3]	刘景琦, 张耀南, 康建芳, 刘杰, 杨治纬, 张智星, 王保得. 基于多智能体协同的高寒山区道路结冰数据工程预警[J]. 数据与计算发展前沿, 2026, 8(2): 40-53.
[4]	姜恩波, 方肖, 秦瑜. FAIR原则与数据密集型科学社区应用实践研究[J]. 数据与计算发展前沿, 2025, 7(3): 111-121.
[5]	王镇远, 田东, 董禹, 乔娜, 单桂华. 多模态交互：从人机协同迈向人智协同[J]. 数据与计算发展前沿, 2025, 7(3): 81-93.
[6]	赵静宜, 陈昕, 周园春. 开放科学监测体系国际经验及启示[J]. 数据与计算发展前沿, 2024, 6(5): 46-56.
[7]	黎建辉, 张丽丽, 陈炜, 刘宇德, 李菁菁, 马彤宇, 罗泽, 张海明, 余宁, 任勇毛, 裴昶华, 李健, 牛铁, 张波, 吴开超, 郑依华. 链接连接联结——中国科技网30年发展回顾与展望[J]. 数据与计算发展前沿, 2024, 6(2): 10-24.
[8]	张丽丽, 陈灿, 张海明, 郑依华, 罗泽, 黎建辉. 互信、互联与互操作：全球开放科学云计划进展与展望[J]. 数据与计算发展前沿, 2024, 6(2): 25-38.
[9]	张丽丽,黎建辉. 科研信息基础设施的运行治理模式研究[J]. 数据与计算发展前沿, 2022, 4(6): 92-104.
[10]	梁思立,姜桂飞,陈泰劼,邓益超,战瑀璠,张玉志. 基于蒙特卡洛树搜索的通用博弈系统的构建与优化研究[J]. 数据与计算发展前沿, 2022, 4(3): 66-77.