随着城市化快速推进和人口迅速增加，城市功能区空间结构发生快速演变。城市功能区合理布局可以促进经济发展、提高土地利用效率、改善居民生活质量，对城市可持续发展及规划管理具有重要意义。本文以Point of Interest（POI）数据和Open Street Map（OSM）路网数据为基础，利用标准差椭圆、平均最邻近、核密度、转移矩阵、信息熵等方法，开展了2013—2023年西安市主城区城市功能区分布特征、演变规律及影响因素的研究，并提出相关优化对策。主要研究结论如下：

（1）2013年，西安市主城区城市功能区主要以混合功能区和商业功能区为主，分别为371.22 km²和215.70 km²，共占研究区总面积的70 %以上。2023年，西安市主城区城市功能区仍以混合功能区和商业功能区为主，但面积扩大，占研究区总面积的93.17 %。2013—2023年，除混合功能区以及商业功能区呈上升趋势外，其他功能区均下降，其中居住功能区以及公共管理与公共服务功能区下降显著。从功能区结构特征上，2013—2023年，信息熵和均衡度均呈现出不断增加的趋势，说明西安市主城区空间结构发生了一定的变化。其中在2013—2018年，信息熵、均衡度的增加，主要表现为居住功能区和绿地与开敞空间功能区面积的增加和混合功能区面积的减少。2018—2023年，信息熵、均衡度的增加，主要表现为工业功能区、交通运输功能区和休闲娱乐功能区面积的增加及商业功能区面积的减少。

（2）2013—2023年西安市主城区城市功能区空间分布特征表明:从聚集程度上看出居住功能区、工业功能区聚集程度升高；商业功能区、公共管理与公共服务功能区、交通运输功能区下降；绿地与开敞空间功能区及休闲娱乐功能区变化不大。从重心转移变化情况可以看出居住功能区、商业功能混合功能区空间转移距离较小，分别转移3.08 km、1.67 km、1.37 km；休闲娱乐功能区和工业功能区均发生了较大改变，分别转移14.55 km和13.96 km；公共管理与公共服务功能区与绿地与开敞空间功能区变化较小。从圈层空间分布可以看出居住功能区逐渐向外围进行迁移；商业功能区分布由较为分散逐渐聚集在距离圈层中心3 km与7 km之间的范围内；工业功能区在距离圈层中心6 km与9 km之间有所增加；交通运输功能区在距离圈层中心7 km与9 km之间分布明显减少，其余各圈层变化不大；混合功能区在2013—2018年减少并向外围转移，2018—2023年逐渐增加并向中心汇集。从转移矩阵结果表明混合功能区向商业功能区转出的面积最大，转出面积达到149.05 km²，向其他功能区转出的面积较小，合计转出2.04 km²；公共管理与公共服务功能区向混合功能区和商业功能区转出面积较大，分别转移48.64 km²和25.50 km²；居住功能区向商业功能区转出的面积最大，转移9.32 km²；交通运输功能区向混合功能区转出面积最大，转出23.56 km²；其余各功能区均转出面积较小且稳定。

（3）从定量和定性的角度分别分析西安市主城区城市功能区时空演变影响因素。通过参数最优地理探测器模型选取高程、坡度、人口密度、人均GDP、路网密度、POI多样性作为西安市主城区城市功能区的影响因子。结果表明POI多样性、人口密度、高程对于城市功能区结构影响力较大，坡度、路网密度、人均GDP影响力相对较小。从自然、交通、社会三个角度定性分析对西安市主城区城市功能区的影响并提出以下优化对策：加强道路系统引导作用；发挥规划政策的调控作用；促进产业布局的优化作用。

With the rapid urbanization and population growth, the spatial structure of urban functional zones has undergone rapid evolution. Rational layout of urban functional zones can promote economic development, improve land use efficiency, and enhance residents' quality of life, which is of great significance to urban sustainable development and planning management. Based on Point of Interest (POI) data and Open Street Map (OSM) road network data, this paper utilizes methods such as standard deviation ellipse, average nearest neighbor, kernel density, transfer matrix, and information entropy to investigate the distribution characteristics, evolution patterns, and influencing factors of urban functional zones in the main urban area of Xi'an from 2013 to 2023. Relevant optimization countermeasures are also proposed. The main research conclusions are as follows:

In 2013, the urban functional zones of Xi'an's main urban district were dominated by Mixed-use and Commercial functional zones, covering 371.22 km² and 215.70 km², respectively, accounting for over 70% of the total study area. By 2023, Mixed-use and Commercial functional zones remained prevalent in Xi'an's main urban district, but their coverage expanded, occupying 93.17% of the total study area. From 2013 to 2023, apart from the increasing trend in Mixed-use and Commercial functional zones, all other functional zones saw a decline, especially Residential functional zones and Public administration and public service functional zones, which dropped significantly. In terms of functional zones structural characteristics, both information entropy and balance degree showed an increasing trend from 2013 to 2023, indicating certain changes in the spatial structure of Xi'an's main urban zones. Specifically, from 2013 to 2018, the increase in information entropy and balance degree was mainly reflected in the expansion of Residential functional zones and Ecological functional zones, coupled with a reduction in Mixed-use functional zones. From 2018 to 2023, the surge in information entropy and balance degree primarily resulted from the growth of Industrial, Transportation, and Recreation district functional zones, accompanied by a decrease in Commercial functional zones.

The spatial distribution characteristics of urban functional zones in the main urban area of Xi'an from 2013 to 2023 indicate the following trends: In terms of aggregation, Residential and Industrial functional zones have become more concentrated, while Commercial,Public administration and public service, and Transportation functional  zones have become less aggregated. Ecological functional and Recreation district functional zones have shown little change. In terms of shifts in the center of gravity, Residential, Commercial, and Mixed-use functional zones have experienced smaller spatial displacements, moving by 3.08 km, 1.67 km, and 1.37 km, respectively. Recreation district and Industrial functional zones have undergone significant changes, shifting by 14.55 km and 13.96 km, respectively. Public administration and public service functional zones and Ecological functional zones have seen minimal changes. Looking at the spatial distribution across concentric circles, Residential functional zones are gradually migrating outwards. Commercial functional zones, which were initially scattered, are increasingly concentrating within a 3 km to 7 km radius from the center. Industrial functional zones have increased between 6 km and 9 km from the center. The distribution of Transportation functional zones between 7 km and 9 km from the center of the sphere decreased significantly, while the other layers changed little. Mixed-use functional zones decreased and shifted outwards from 2013 to 2018 but gradually increased and converged towards the center from 2018 to 2023. Transfer matrix results show that Mixed-use functional zones has the largest outgoing transfer to the Commercial functional zones, with a transferred area of 149.05 km². Transfers to other functional zones are smaller, totaling 2.04 km². Public administration and public service functional zones has a significant outgoing transfer to Mixed-use and Commercial functional zones, transferring 48.64 km² and 25.50 km², respectively. Residential functional zones has the largest outgoing transfer to the Commercial functional zones, transferring 9.32 km². Transportation functional zones has the largest outgoing transfer to Mixed-use functional zones, transferring 23.56 km². Other functional zones have smaller and more stable outgoing transfers

This paper quantitatively and qualitatively analyzes the influencing factors of spatio-temporal evolution of urban functional areas in the main city zone of Xi'an. By adopting the optimal parameter geographical detector model, elevation, slope, population density, GDP per capita, road network density, and POI diversity are chosen as the influencing factors of urban functional zones in the main city zone of Xi'an. The results suggest that POI diversity, population density, and elevation exert a greater influence on the structure of urban functional zones, whereas slope, road network density, and GDP per capita have relatively minor effects. From natural, transportation, and societal perspectives, a qualitative analysis is conducted on the impact of urban functional areas in the main city zone of Xi'an, and the following optimization countermeasures are proposed: strengthening the guiding role of the road system, giving full play to the regulatory effect of planning policies, and promoting the optimal allocation of industrial layout.

[1]汪光焘, 叶青, 李芬, 等. 培育现代化都市圈的若干思考[J]. 城市规划学刊, 2019, 15(5): 14-23.

[2]张可云, 戴美卉, 王洋志. 中国城市型政区调整与城市化的适应性分析[J]. 地理学报, 2023, 78(12): 3129-3143.

[3]郝宪印, 张念明. 新时代我国区域发展战略的演化脉络与推进路径[J]. 管理世界, 2023, 39(1): 56-68.

[4]童陆亿. 国内外城市扩张内涵及度量研究进展[J]. 世界地理研究, 2020, 29(4): 762-772.

[5]Jiang P H, Cheng Q W, Gong Y, et al. Using Urban Development Boundaries to Constrain Uncontrolled Urban Sprawl in China[J]. Annals of the American Association of Geographers, 2016, 106(6): 1321-1343.

[6]朱伟. 不确定世界的二元性治理： 构建常态与应急的平衡[J]. 江西社会科学, 2020, 40(12): 199-209.

[7]马志远, 段学军, 王磊, 等. 长江经济带区域发展与资源环境承载力空间耦合特征及高质量发展路径[J]. 长江流域资源与环境, 2022, 31(9): 1873-1883.

[8]张运平, 林建平, 黄艺敏, 等. 基于PLUS空间决策模型的城镇与农业空间协调布局优化[J]. 农业工程学报, 2023, 39(20): 233-243.

[9]焦雅楠, 杜鹏, 刘万波, 等. 烟台城市扩展特征及驱动力分析[J]. 云南地理环境研究, 2022, 34(6): 26-33.

[10]西安市规划局, 《西安城市总体规划(2008年—2020年)》概要[J]. 建筑与文化, 2008, (7): 9-17.

[11]李梦洁, 姜翠梅. 国土空间背景下乡村规划体系构建——基于西安市的探索与经验[J]. 小城镇建设, 2023, 41(9): 76-82.

[12]范晓鹏, 郄海潮. 西安都市圈中心城区的时空演变及影响机制研究[J]. 城市发展研究, 2021, 28(6): 34-39.

[13]Su C, Hu X L, Meng Q Y, et al. A multimodal fusion framework for urban scene understanding and functional identification using geospatial data[J]. International Journal of Applied Earth Observation and Geoinformation, 2024, 127: 103696.

[14]Chen Y Y, Zhao P J, Lin Y, et al. Semantic-Enhanced Graph Convolutional Neural Networks for Multi-Scale Urban Functional-Feature Identification Based on Human Mobility[J]. ISPRS International Journal of Geo-Information, 2024, 13(1): 27-49.

[15]Luo G W, Ye J Y, Wang J F, et al. Urban Functional Zone Classification Based on POI Data and Machine Learning[J]. Sustainability, 2023, 15(5): 4631-4648.

[16]Hu J J, Gao Y, Xue C, et al. Recognizing mixed urban functions from human activities using representation learning methods[J]. International Journal of Digital Earth, 2023, 16(1): 289-307.

[17]Yu M Y, Xu H Q, Zhou F L, et al. A Deep-Learning-Based Multimodal Data Fusion Framework for Urban Region Function Recognition[J]. ISPRS International Journal of Geo-Information, 2023, 12(12): 468.

[18]Zhang J F, Tao J Y, Xu L, et al. RETRACTED: Recognition and zoning optimization of geographical spatial conflict in Wuhan metropolitan area based on multi-functional perspective[J]. International Journal of Electrical Engineering & Education, 2023, 60(1): 2850-2870.

[19]郑国强, 乔宇昊, 孙思民, 等. 基于POI数据的城市功能区识别研究[J]. 地理空间信息, 2023, 21(10): 58-61.

[20]罗婷, 兰强. 超大城市的功能混合度识别与区域协调发展研究——以成都市为例[C]//中国城市规划学会. 人民城市, 规划赋能——2023中国城市规划年会论文集（14区域规划与城市经济）. 重庆: 重庆大学出版社, 2023: 79-88.

[21]曹元晖, 刘纪平, 王勇, 等.基于POI数据的城市建筑功能分类方法研究[J]. 地球信息科学学报, 2020, 22(6): 1339-1348.

[22]Li Y P, Zhao J S, Zhang S Q, et al. Qualitative-quantitative identification and functional zoning analysis of production-living-ecological space: a case study of Urban Agglomeration in Central Yunnan, China.[J]. Environmental monitoring and assessment, 2023, 195(10): 1163.

[23]Liu T, Cheng G, Yang J. Multi-Scale Recursive Identification of Urban Functional zones Based on Multi-Source Data[J]. Sustainability, 2023, 15(18): 13870.

[24]Chen P, Sheng Z W, EnLong L, et al. Identification of urban functional zones in Macau Peninsula based on POI data and remote information sensors technology for sustainable development[J]. Physics and Chemistry of the Earth, 2023, 131: 103447.

[25]Siti I H N, Virginia S, F. W R C. Functional urban ground-cover plants: identifying traits that promote rainwater retention and dissipation[J]. Urban Ecosystems, 2023, 26(6): 1709-1724.

[26]Yang Y Y, Ren X Z, Yang J M. Trade-offs or synergies? Identifying dynamic land use functions and their interrelations at the grid scale in urban agglomeration[J]. Cities, 2023, 140: 104384.

[27]韩怡然. 基于大数据的城市功能空间识别与演变分析[D]. 郑州: 河南财经政法大学, 2023.

[28]Sun Z H, Li P, Wang D C, et al. Recognizing Urban Functional Zones by GF-7 Satellite Stereo Imagery and POI Data[J]. Applied Sciences, 2023, 13(10): 6300.

[29]Wang J X, Gao C R, Wang M M, et al. Identification of Urban Functional zones and Urban Spatial Structure Analysis by Fusing Multi-Source Data Features: A Case Study of Zhengzhou, China[J]. Sustainability, 2023, 15(8): 1-23.

[30]Giacomo C, Matteo B, Chiara L, et al. Automatic identification of urban functions via social mining[J]. Cities, 2023, 137: 104262.

[31]李瑞琦. 吉林市城市空间形态演变与功能区识别研究[D]. 长春: 吉林大学, 2023.

[32]段彬. 基于POI数据的贵阳市功能区识别与时空分析[D]. 贵阳: 贵州大学, 2023.

[33]Zhang Y ,Liu P Y, Filip B. Knowledge and topology: A two layer spatially dependent graph neural networks to identify urban functions with time-series street view image[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2023, 198: 153-168.

[34]李景琪. 基于POI数据的山东省滨海城市功能区时空格局演变分析[D]. 济南: 山东建筑大学, 2022.

[35]Wang Z S, Lu F D, Liu Z H, et al. Measuring spatial nonstationary effects of POI-based mixed use on urban vibrancy using Bayesian spatially varying coefficients model[J]. International Journal of Geographical Information Science, 2023, 37(2): 339-359.

[36]薛冰, 赵冰玉, 李京忠, 等. 地理大数据中POI数据质量的评估与提升方法[J]. 地理学报, 2023, 78(5): 1290-1303.

[37]曾穗平,吕艳梅, 田健, 等. 智能算法在城市形态优化研究中的演化路径与应用情景——基于Citespace知识图谱的分析[J]. 城市问题, 2022, (4): 14-23.

[38]武新宇, 孙立双, 谢志伟, 等. 遥感影像与POI数据相结合的城市建成区提取适用性研究——以沈阳市为例[J]. 测绘通报, 2019, (11): 134-136+159.

[39]王俊秀. 公交出行视角下山地城市功能区空间交互特征研究[D]. 重庆: 重庆交通大学, 2022.

[40]Huang J W, Chu C Q, Wang L, et al. Research on the Extraction Method Comparison and Spatial-Temporal Pattern Evolution for the Built-Up Area of Hefei Based on Multi-Source Data Fusion[J]. Remote Sensing, 2023, 15(23): 5617.

[41]陈相利. 非首都功能疏解下北京城市空间结构演变及疏解腾退资源分布识别[D]. 北京: 中国地质大学(北京), 2020.

[42]陶煜, 刘宛. 基于“评估—优化”的历史城区小微公共空间更新方法——以北京什刹海历史文化街区为例[J]. 规划师, 2024, 40(1): 91-99.

[43]邵莉婷, 邵大伟, 吴殿鸣, 等. 公园绿地与居住用地的空间耦合模式及影响因素——以南京为例[J]. 中国园林, 2023, 39(8): 90-95.

[44]苟爱萍, 张振, 王江波, 等. 基于POI视角的上海城市功能区演化特征及驱动因素[J]. 资源开发与市场, 2022, 38(9): 1052-1062.

[45]熊晓轶, 王蒙蒙. 协同治理视角下京津冀地区流通产业同质化的演变逻辑[J]. 商业经济研究, 2022, (8): 166-169.

[46]刘威, 雷楠, 刘大齐, 等. 太原市住宅区与零售业关联分析——POI应用示例[J]. 科技创新与应用, 2024, 14(1): 123-128.

[47]张慧馨. 基于POI数据的开封市复合功能格局演变与优化策略[D]. 开封: 河南大学, 2023.

[48]郑晓伟. 基于开放数据的西安城市中心体系识别与优化[J]. 规划师, 2017, 33(1): 57-64.

[49]窦旺胜, 王成新, 薛明月, 等. 基于POI数据的城市用地功能识别与评价研究——以济南市内五区为例[J]. 世界地理研究, 2020, 29(4): 804-813.

[50]薛冰, 肖骁, 李京忠, 等. 基于兴趣点（POI）大数据的东北城市空间结构分析[J]. 地理科学, 2020, 40(5): 691-700.

[51]张汝华, 郭森垚, 施庆利. 基于信息熵的山东省货运结构演变及动因分析[J]. 重庆交通大学学报(自然科学版), 2022, 41(12): 33-40.

[52]程辉, 张延军, 吴旭影, 等. 基于ArcGIS与信息熵法的吉林省城市用地变化分析[J]. 中国软科学, 2020, (S1): 181-186.

[53]张军学, 范广忠, 陈哲衡, 等. 耦合信息熵模式下的城市绿地设计可持续能值研究[J]. 城市建筑, 2023, 20(12): 15-17.

[54]杨月恒, 郭力娜, 姜广辉, 等. 城市三生空间功能格局与分异特征分析-以辽宁省为例[J]. 华北理工大学学报(自然科学版), 2022, 44(3): 94-102.

[55]李锋, 王如松. 城市绿地系统的生态服务功能评价、规划与预测研究——以扬州市为例[J]. 生态学报, 2003, 23(9): 1929-1936.

[56]李玉娟, 彭仕兰. 基于标准差椭圆的长三角地区产业集聚特征分析[J]. 统计与决策, 2024, 40(1): 136-141.

[57]万红莲, 王晓利, 黄敏, 等. 基于POI数据的关中平原城市群县域旅游要素空间格局及影响因素研究[J]. 干旱区资源与环境, 2024, 38(6): 200-208.

[58]骆元家, 梁庆璇, 刘洪, 等. 1990—2020年南宁市土地利用时空变化[J]. 湖北农业科学, 2024, 63(2): 128-136.

[59]王劲峰, 徐成东. 地理探测器: 原理与展望[J]. 地理学报, 2017, 72(1): 116-134.

[60]胡晓鸣, 黎小龙, 蔚芳, 等. 基于POI的城市功能区及其混合度识别研究——以重庆市核心城区为例[J]. 西南大学学报(自然科学版), 2021, 43(1): 164-173.

[61]赵卫锋, 李清泉, 李必军, 等. 利用城市POI数据提取分层地标[J]. 遥感学报, 2011, 15(5): 973-988.

[62]谢孟, 程怡诺, 范胜龙, 等. 基于POI数据的福州城市功能区识别[J]. 农业与技术, 2023, 43(13): 107-112.

[63]贺泽宇, 年雁云, 陈思文, 等. 基于出租车数据的城市功能区分析——以西安市主城区为例[J]. 城市勘测, 2022, 192(4): 17-22.

[64]李鹏, 王英杰, 虞虎, 等. 基于GIS格网化分析支撑的旅游空间规划技术方法研究——以青岛市为例[J]. 自然资源学报, 2018, 33(5): 813-827.

[65]陈占龙, 周路林, 禹文豪, 等. 顾及兴趣点潜在上下文关系的城市功能区识别[J]. 测绘学报, 2020, 49(7): 907-920.

[66]曹元晖, 刘纪平, 王勇, 等. 基于POI数据的城市建筑功能分类方法研究[J]. 地球信息科学学报, 2020, 22(6): 1339-1348.

[67]贾斐雪, 闫金凤, 王甜, 等. 大数据构建的赋分评价模型与功能区识别研究[J]. 测绘科学, 2021, 46(8): 172-178.

[68]杨鑫, 杨宜来, 胡克勇, 等. 基于POI和主动学习的城市功能区分类方法研究[J]. 河北省科学院学报, 2022, 39(2): 23-33.

[69]周恺, 张旭阳, 夏依宁, 等. 人口—土地协调视角下的湖南省主体功能区优化探索[J]. 资源科学, 2024, 46(2): 421-433.

[70]陈小利, 周航. “高质量、共富裕”背景下，城市融合区的交通发展思路——以杭州大城北地区为例[C]//中国城市规划学会城市交通规划学术委员会. 绿色·智慧·融合——2021/2022年中国城市交通规划年会论文集. 杭州: 杭州市规划设计研究院, 2022: 652-666.

[71]刘薇, 常振海. 关中平原城市群“一圈一轴三带”空间规划目标进展分析[J]. 天水师范学院学报, 2023, 43(2): 118-128.

[72]王登海. 优化产业链布局西安都市圈建设提速[N]. 中国经营报, 2023-07-17(B09).