MapColor-Agent：融合大语言模型与多智能体的行政区划图配色框架

doi:10.11947/j.AGCS.2026.20250436

测绘学报 ›› 2026, Vol. 55 ›› Issue (2): 359-370.doi: 10.11947/j.AGCS.2026.20250436

• 地图学与地理信息 • 上一篇

MapColor-Agent：融合大语言模型与多智能体的行政区划图配色框架

魏智威¹^,²(), 杨乃³(), 王一杰³, 廖华¹^,², 周梦杰¹^,², 许文嘉⁴

^1.湖南师范大学地理科学学院，湖南　长沙　410081
^2.地理空间大数据挖掘与应用湖南省重点实验室，湖南　长沙　410081
^3.中国地质大学（武汉）地理与信息工程学院，湖北　武汉　430074
^4.北京邮电大学信息与通信工程学院，北京　100876

收稿日期:2025-10-20 修回日期:2026-01-07 发布日期:2026-03-13
通讯作者: 杨乃 E-mail:trentonwei@whu.edu.cn;yangnai@cug.edu.cn
作者简介:魏智威（1993—），男，博士，讲师，研究方向为地理可视化与地理大模型应用。 E-mail：trentonwei@whu.edu.cn
基金资助:
国家自然科学基金(42501551; 42371455; 42171438)

MapColor-Agent：a large language model-integrated multi-agent framework for administrative map color design

Zhiwei WEI¹^,²(), Nai YANG³(), Yijie WANG³, Hua LIAO¹^,², Mengjie ZHOU¹^,², Wenjia XU⁴

^1.School of Geographic Sciences, Hunan Normal University, Changsha 410081, China
^2.Hunan Key Laboratory of Geospatial Big Data Mining and Application, Changsha 410081, China
^3.School of Geography and Information Engineering, China University of Geosciences, Wuhan 430074, China
^4.School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China

Received:2025-10-20 Revised:2026-01-07 Published:2026-03-13
Contact: Nai YANG E-mail:trentonwei@whu.edu.cn;yangnai@cug.edu.cn
About author:WEI Zhiwei (1993—), male, PhD, lecturer, majors in Geo-VIS and Geo-FM based applications. E-mail: trentonwei@whu.edu.cn
Supported by:
The National Natural Science Foundation of China(42501551; 42371455; 42171438)

摘要/Abstract

摘要：

近年来，大语言模型在语义理解与任务推理方面取得显著进展，为地图设计智能化提供了一种技术途径。本文以行政区划图的配色设计为研究对象，提出并构建了一个融合大语言模型与多智能体的交互式智能地图配色框架——MapColor-Agent。该系统以大语言模型为语义理解核心，通过多智能体实现任务分解与过程协同，并结合自然语言和图形化操作界面，支持用户以更直观的方式生成符合需求的地图配色方案。本文采用系统可用性量表与半结构化访谈评估系统性能。结果表明，MapColor-Agent的总体可用性得分为77.9，达到良好水平。用户普遍认为系统学习成本低、操作流程清晰、交互体验自然，且具有较高的可理解性和控制性。差异性分析显示，熟悉地图配色的用户在学习效率与系统复杂度感知方面得分更高，说明知识背景对系统使用效果具有一定影响。访谈结果进一步表明，系统在语义理解与任务引导方面表现出优势，但在复杂语义解析与生成稳定性方面仍有改进空间。研究结果验证了大语言模型与智能体协同架构在地图配色设计中的可行性，可为未来基于语义推理与多模态交互的地图智能设计提供参考。

关键词: 大语言模型, 多智能体系统, 人机交互, 地图配色设计, 行政区划图

Abstract:

In recent years, large language models (LLMs) have made remarkable progress in semantic understanding and task reasoning, providing new technological opportunities for the intelligent design of maps. This study focuses on the color design of administrative maps and proposes a framework, MapColor-Agent, that integrates large language models with a multi-agent collaboration mechanism. The framework employs the LLMs as the semantic reasoning core and enables task decomposition and process coordination through multiple agents. It also combines natural language interaction with a graphical user interface, allowing users to intuitively generate semantically consistent color schemes for maps. The system performance was evaluated using the system usability scale and semi-structured interviews. The results show that MapColor-Agent achieved an overall usability score of 77.9, indicating a good level of usability. Participants generally found the system easy to learn, clear in operation, and natural in interaction, with high levels of interpretability and controllability of results. Difference analysis revealed that participants familiar with map color design scored higher in learning efficiency and perceived complexity, suggesting that background knowledge influences user experience. The interview results further indicated that the framework performs well in semantic understanding and task guidance, though improvements are needed in complex semantic parsing and generation stability. Overall, the findings demonstrate the feasibility of integrating large language models with multi-agent collaboration for map color design and provide a reference for future research on semantic reasoning and multimodal interaction in intelligent cartographic design.

Key words: large language models, multi-agent system, human-computer interaction, map color design, administrative map

中图分类号:

P283

魏智威, 杨乃, 王一杰, 廖华, 周梦杰, 许文嘉. MapColor-Agent：融合大语言模型与多智能体的行政区划图配色框架[J]. 测绘学报, 2026, 55(2): 359-370.

Zhiwei WEI, Nai YANG, Yijie WANG, Hua LIAO, Mengjie ZHOU, Wenjia XU. MapColor-Agent：a large language model-integrated multi-agent framework for administrative map color design[J]. Acta Geodaetica et Cartographica Sinica, 2026, 55(2): 359-370.

图/表 4

图1

图2

图3

表1

参考文献 25

[1]	苏世亮, 向娟, 杜清运, 等. 叙事地图学的概念框架构建[J]. 测绘学报, 2025, 54(8): 1371-1388. DOI: . doi: 10.11947/j.AGCS.2025.20250107
	SU Shiliang, XIANG Juan, DU Qingyun, et al. Re-conceptualizing narrative cartography[J]. Acta Geodaetica et Cartographica Sinica, 2025, 54(8): 1371-1388. DOI: . doi: 10.11947/j.AGCS.2025.20250107
[2]	吴明光, 乔莉鸽, 闾国年. 从视觉神经科学的视角看地图视觉变量理论[J]. 地理学报, 2024, 79(9): 2191-2205.
	WU Mingguang, QIAO Lige, LÜ Guonian. Revisiting cartographic visual variables from a perspective of visual neuroscience[J]. Acta Geographica Sinica, 2024, 79(9): 2191-2205.
[3]	吴明光, 孙彦杰, 闾国年. 地图风格迁移：概念、综述与展望[J]. 武汉大学学报(信息科学版), 2022, 47(12): 2069-2084.
	WU Mingguang, SUN Yanjie, LÜ Guonian. Cartographic style transfer: idea, review and envision[J]. Geomatics and Information Science of Wuhan University, 2022, 47(12): 2069-2084.
[4]	魏智威, 郭庆胜, 颜芬, 等. 顾及视觉规律的行政区划图自动着色回溯法[J]. 测绘学报, 2018, 47(3): 396-402. DOI: . doi: 10.11947/j.AGCS.2018.20170148
	WEI Zhiwei, GUO Qingsheng, YAN Fen, et al. Backtracking method of coloring administrative maps considering visual perception rules[J]. Acta Geodaetica et Cartographica Sinica, 2018, 47(3): 396-402. DOI: . doi: 10.11947/j.AGCS.2018.20170148
[5]	高建军, 关焱, 王友亮. 专题地图基调色设计研究[J]. 武汉大学学报(信息科学版), 2008, 33(7): 766-769.
	GAO Jianjun, GUAN Yan, WANG Youliang. Design of basic tone of thematic map color[J]. Geomatics and Information Science of Wunan University, 2008, 33(7): 766-769.
[6]	韩云, 郭庆胜, 章莉萍, 等. 行政区划图自动着色的混合遗传算法[J]. 武汉大学学报(信息科学版), 2007, 32(8): 748-751.
	HAN Yun, GUO Qingsheng, ZHANG Liping, et al. Hybrid genetic algorithm for coloring administrative map[J]. Geomatics and Information Science of Wunan University, 2007, 32(8): 748-751.
[7]	彭珊鸰, 何宗宜, 宋鹰, 等. 图着色的混合遗传算法研究[J]. 武汉大学学报(信息科学版), 2009, 34(12): 1499-1501.
	PENG Shanling, HE Zongyi, SONG Ying, et al. A hybrid genetic algorithm for coloring map[J]. Geomatics and Information Science of Wuhan University, 2009, 34(12): 1499-1501.
[8]	STEINRÜCKEN J, PLÜMER L. Identification of optimal colours for maps from the web[J]. The Cartographic Journal, 2013, 50(1): 19-32.
[9]	BREWER C A. Prediction of simultaneous contrast between map colors with Hunt's model of color appearance[J]. Color Research & Application, 1996, 21(3): 221-235.
[10]	CHESNEAU E. A model for the automatic improvement of colour contrasts in maps: application to risk maps[J]. International Journal of Geographical Information Science, 2011, 25(1): 89-111.
[11]	BREWER C A. The development of process-printed munsell charts for selecting map colors[J]. The American Cartographer, 1989, 16(4): 269-278.
[12]	BREWER C A. Spectral schemes: controversial color use on maps[J]. Cartography and Geographic Information Systems, 1997, 24(4): 203-220.
[13]	SUCHAN T A, BREWER C A. Qualitative methods for research on mapmaking and map use[J]. The Professional Geographer, 2000, 52(1): 145-154.
[14]	BREWER C A, PICKLE L. Evaluation of methods for classifying epidemiological data on choropleth maps in series[J]. Annals of the Association of American Geographers, 2002, 92(4): 662-681.
[15]	HARROWER M, BREWER C A. ColorBrewer.org: an online tool for selecting colour schemes for maps[J]. The Cartographic Journal, 2003, 40(1): 27-37.
[16]	BREWER C A. Basic mapping principles for visualizing cancer data using geographic information systems (GIS)[J]. American Journal of Preventive Medicine, 2006, 30(2): S25-S36.
[17]	HOU Yihan, YANG Manling, CUI Hao, et al. C2Ideas: supporting creative interior color design ideation with a large language model[C]//Proceedings of 2024 CHI Conference on Human Factors in Computing Systems. New York: ACM Press, 2024: 172.
[18]	YANG Nai, WANG Yijie, WEI Zhiwei, et al. MapColorAI: designing contextually relevant choropleth map color schemes using a large language model[J/OL]. Cartography and Geographic Information Science. [2025-09-20]. https://doi.org/10.1080/15230406.2025.2531055.
[19]	ZHANG Yifan, ZHANG Wenbo, ZENG Ziyi, et al. MapReader: a framework for learning a visual language model for map analysis[J]. International Journal of Geographical Information Science, 2025, 39(9): 1973-2008.
[20]	苏世亮, 李子纯, 杜清运, 等. 叙事地图的符号体系及叙述机制[J]. 测绘学报, 2025, 54(1): 165-181. DOI: . doi: 10.11947/j.AGCS.2025.20240175
	SU Shiliang, LI Zichun, DU Qingyun, et al. Symbols of narrative maps: compositional structure and working mechanism[J]. Acta Geodaetica et Cartographica Sinica, 2025, 54(1): 165-181. DOI: . doi: 10.11947/j.AGCS.2025.20240175
[21]	李千千, 苏世亮. 数智媒介下叙事地图的基本理论问题——概念界定、叙事结构与意义生成机制[J/OL]. 武汉大学学报(信息科学版). [2025-09-20]. https://doi.org/10.13203/j.whugis20250261.
	LI Qianqian, SU Shiliang. Fundamental theoretical issues of narrative maps under digital-intelligent media: definition of concepts, narrative structure, and meaning-generating mechanism[J/OL]. Geomatics and Information Science of Wuhan University. [2025-09-20]. https://doi.org/10.13203/j.whugis20250261.
[22]	何宗宜, 宋鹰. 普通地图编制[M]. 武汉: 武汉大学出版社, 2015: 61-78.
	HE Zongyi, SONG Ying. General map compilation[M]. Wuhan: Wuhan University Press, 2015: 61-78.
[23]	GUAN Yanchu, WANG Dong, CHU Zhixuan, et al. Intelligent agents with LLM-based process automation[C]//Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York: ACM Press, 2024: 5018-5027.
[24]	WEI J, WANG Xuezhi, SCHUURMANS D, et al. Chain-of-thought prompting elicits reasoning in large language models[EB/OL]. [2025-09-20]. https://arxiv.org/pdf/2201.11903.
[25]	LEWIS J R. The system usability scale: past, present, and future[J]. International Journal of Human-Computer Interaction, 2018, 34(7): 577-590.

问题	影响因素	平均值	方差齐性检验（p值）	独立样本T检验（p值）	Welch's T检验（p值）	Cronbach's α值
Q1	不熟悉	4.16	0.714	0.955	—	0.016
Q1	熟悉	4.18	0.714	0.955	—	0.016
Q2	不熟悉	2.14	0.838	0.031^**	—	0.634
Q2	熟悉	1.65	0.838	0.031^**	—	0.634
Q3	不熟悉	4.12	0.562	0.010^**	—	0.760
Q3	熟悉	4.65	0.562	0.010^**	—	0.760
Q4	不熟悉	2.28	0.087^*	0.002^***	—	0.940
Q4	熟悉	1.47	0.087^*	0.002^***	—	0.940
Q5	不熟悉	3.88	0.926	0.424	—	0.231
Q5	熟悉	4.06	0.926	0.424	—	0.231
Q6	不熟悉	2.14	0.443	0.441	—	0.222
Q6	熟悉	1.94	0.443	0.441	—	0.222
Q7	不熟悉	4.49	0.012^**	—	0.052^**	0.508
Q7	熟悉	4.82	0.012^**	—	0.052^**	0.508
Q8	不熟悉	1.70	0.756	0.161	—	0.434
Q8	熟悉	1.41	0.756	0.161	—	0.434
Q9	不熟悉	3.93	0.201	0.654	—	0.159
Q9	熟悉	4.06	0.201	0.654	—	0.159
Q10	不熟悉	2.00	0.136	0.812	—	0.068
Q10	熟悉	1.94	0.136	0.812	—	0.068
总得分	不熟悉	75.81	0.994	0.036^**	—	0.615
总得分	熟悉	83.38	0.994	0.036^**	—	0.615

MapColor-Agent：融合大语言模型与多智能体的行政区划图配色框架

MapColor-Agent：a large language model-integrated multi-agent framework for administrative map color design

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 4

参考文献 25

相关文章 5

编辑推荐

Metrics

本文评价

[1]	李维炼, 冉晴晴, 党沛, 朱军, 朱庆, 张恒. 耦合LLMs-KG的地下车站设施洪水脆弱性级联效应分析方法[J]. 测绘学报, 2026, 55(1): 154-168.
[2]	吴华意, 沈张骁, 侯树洋, 梁健源, 赵安琪, 矫皓月, 桂志鹏, 关雪峰. 大语言模型驱动的GIS分析：方法、应用与展望[J]. 测绘学报, 2025, 54(4): 621-635.
[3]	胡鑫, 杨学习, 江一凡, 王宪彬, 丁晨, 谢顾然, 邓敏. 基于多智能体层次化协同的地理事件抽取与时空解析[J]. 测绘学报, 2025, 54(11): 2052-2067.
[4]	汤圣君, 杜思齐, 王伟玺, 郭仁忠. 面向室内空间智能的三维场景图表达与应用[J]. 测绘学报, 2024, 53(7): 1355-1370.
[5]	魏智威, 郭庆胜, 颜芬, 王勇. 顾及视觉规律的行政区划图自动着色回溯法[J]. 测绘学报, 2018, 47(3): 396-402.