<p>随着城市化进程的加速，地铁作为一种高效的出行方式，面临着日益严峻的应急疏散挑战。特殊人群（老年人、残疾人、孕妇及儿童）因行动能力受限，在疏散中往往面临更高风险。本文聚焦地铁站台应急疏散场景，综合运用文献分析、问卷调查与仿真模拟方法，以西安大雁塔地铁站为研究对象，构建差异化人群疏散模型，探究不同时段特殊人群的疏散效率，研究不同引导员数量策略和不同扶梯开停策略下的疏散效率与瓶颈问题，并提出针对性优化策略。主要内容如下：</p> <p>（1）基于多维度客流调查与行为分析，量化了特殊人群的时空分布特征。现场人员调查发现，节假日特殊人群比例增至29%，老年人移动速度仅为正常人群的60%~70%。问卷调查显示，64.2%的乘客缺乏扶梯应急功能认知，90.4%的人员需要引导员帮助，探究了特殊人群基本特征、安全意识、应急疏散行为等及其影响因素，为模型构建提供了关键行为参数。</p> <p>（2）结合AnyLogic平台构建三维动态疏散仿真模型，揭示了不同场景下的疏散效率差异。结果表明：低峰期因老年人群比例高导致疏散效率减少10%；节假日工况下特殊人群占比增至29%，总疏散时间延长8%；引导员增至5人时，疏散时间缩短10.2%，验证了人工干预与设施调控的协同效应；四种扶梯开停策略（上行运行/下行停止、双扶梯停止、仅楼梯疏散、下行扶梯逆行）中，上行运行/下行停止策略综合效率最高，仿真模型疏散时间为303.2 s，数学模型疏散时间为295.6 s，误差仅为2.51%，可作为平衡方案适用于各种场景；逆行策略虽然疏散效率最高，疏散时间为295 s，但需解决设备逆向运行的技术难题。所有疏散策略误差均控制在3%以内，验证了仿真模型与数学模型的科学一致性，体现了模型的可靠性。</p> <p>（3）基于设立引导员和不同扶梯开停策略仿真数据分析结果，针对应急疏散的瓶颈区域提出应急疏散系统优化方案，通过多策略对比与瓶颈识别，提出关键优化方案：增设2台宽体闸机使出口滞留时间减少55.6%；楼梯宽度增至3.5 m可降低密度峰值40%；设置3 m导流栏杆使路径冲突频率下降67%，并通过现场验证证明优化导流栏杆方案的可行性与实用性。每种优化方案对提高疏散效率都起到了一定的作用，楼梯宽度优化是系统性提升疏散效率的核心措施；导流栏杆优化对特殊人群保护效果显著，建议优先采用；闸机优化可作为低成本补充方案，但需与其他措施配合使用。</p> <p>本文将设立引导员策略和扶梯开停策略与特殊人群疏散行为相结合，提出了针对性的优化方案，为包容性交通系统设计提供理论依据与实践参考，为地铁站应急疏散管理提供了切实可行的建议，对提升城市公共交通安全性与包容性具有重要现实意义。</p>

<p>With the acceleration of urbanization, the subway, as an efficient mode of transportation, is facing increasingly severe challenges in emergency evacuation. Special groups (such as the elderly, the disabled, pregnant women, and children) are at higher risk due to their limited mobility during evacuation. This paper focuses on the emergency evacuation scenarios at subway platforms, comprehensively applying literature analysis, questionnaire surveys, and simulation methods. Taking Xi&#39;an Dayanta subway station as the research object, a differentiated crowd evacuation model is constructed to explore the evacuation efficiency of special groups at different times, study the evacuation efficiency and bottleneck problems under different numbers of guides and different escalator operation strategies, and propose targeted optimization strategies. The main contents are as follows:</p> <p>(1) Based on multi-dimensional passenger flow surveys and behavioral analysis, the spatio-temporal distribution characteristics of special groups are quantified. On-site personnel surveys found that the proportion of special groups increased to 29% during holidays, and the movement speed of the elderly was only 60% to 70% of that of the normal population. Questionnaire surveys showed that 64.2% of passengers lacked awareness of the emergency functions of escalators, and 90.4% needed the help of guides. The basic characteristics, safety awareness, and emergency evacuation behaviors of special groups and their influencing factors were explored, providing key behavioral parameters for model construction.</p> <p>(2) A three-dimensional dynamic evacuation simulation model was constructed using the AnyLogic platform, revealing the differences in evacuation efficiency under different scenarios. The results showed that during off-peak hours, the evacuation efficiency was reduced by 10% due to the high proportion of the elderly; during holiday conditions, the proportion of special groups increased to 29%, and the total evacuation time was extended by 8%; when the number of guides increased to 5, the evacuation time was shortened by 10.2%, verifying the synergy of manual intervention and facility control. Among the four escalator operation strategies (upward operation/downward stop, both escalators stop, only stairs evacuation, and downward escalator reverse operation), the upward operation/downward stop strategy had the highest overall efficiency, with a simulation model evacuation time of 303.2 seconds and a mathematical model evacuation time of 295.6 seconds, with an error of only 2.51%, which can be used as a balanced solution for various scenarios. Although the reverse operation strategy had the highest evacuation efficiency, with an evacuation time of 295 seconds, it needed to solve the technical problem of reverse operation of the equipment. The error of all strategies was controlled within 3%, verifying the scientific consistency of the simulation model and the mathematical model and demonstrating the reliability of the model.</p> <p>(3) Based on the simulation data analysis of the guide setup and different escalator operation strategies, an emergency evacuation system optimization plan was proposed for the core bottleneck areas. Through multi-strategy comparison and bottleneck identification, key optimization plans were proposed: adding 2 wide-body gates reduced the exit retention time by 55.6%; increasing the stair width to 3.5 meters reduced the density peak by 40%; setting 3-meter diversion barriers reduced the path conflict frequency by 67%&nbsp;and through field verification, the feasibility and practicability of optimizing the diversion railing were proved. Each optimization plan played a certain role in improving the evacuation efficiency. The optimization of stair width was the core measure for systematically improving the evacuation efficiency; the optimization of diversion barriers had a significant protective effect on special groups and was recommended for priority adoption; the optimization of gates could be used as a low-cost supplementary solution but needed to be used in combination with other measures.</p> <p>This paper combines the guide setup strategy and escalator operation strategy with the evacuation behavior of special groups, proposes targeted optimization plans, provides theoretical basis and practical reference for the design of inclusive transportation systems, and offers practical suggestions for emergency evacuation management at subway stations. It has important practical significance for improving the safety and inclusiveness of urban public transportation.</p>

[1]曹向辉. 地铁火灾人员疏散影响因素分析及疏散策略 [J]. 中国人民警察大学学报, 2023, 39(8): 48-51+57.

[2]张野, 单妥, 王晓磊, 等. 基于BIM的地铁车站火灾疏散应用研究 [J]. 消防界(电子版), 2023, 9(15): 96-98.

[3]齐若星. 地铁站突发火灾事故中乘客疏散行为决策的影响因素及提升策略研究 [D]. 西安: 西安建筑科技大学, 2024.

[4]刘鑫. 地铁站楼扶梯不同运行工况下乘客应急疏散模型及优化 [D]. 西安：西安科技大学, 2022.

[5]中华人民共和国住房和城乡建设部.地铁设计防火标准: GB 51298-2018 [S]. 中国计划出版社, 2018.

[6]赵艳. 建筑火灾应急疏散与救援技术研究 [C]//中国消防协会学术工作委员会消防科技论文集(2024)—建筑火灾防治技术. 河北省廊坊市安次区消防救援大队, 2024: 81-83.

[7]张克诚. 火灾场景下地铁站台乘客疏散时间的预测方法分析 [J]. 电工技术, 2024, (20): 175-177+181.

[8]董利辉, 周海竹, 邓云峰. 地铁火灾非适应性疏散的从众行为影响因素分析 [J]. 中国安全生产科学技术, 2024, 20(7): 108-113.

[9]付君窈. 地铁车站突发事件行人疏散行为研究 [J]. 交通科技与管理, 2024, 5(13): 14-17.

[10]金佳宁. 大型高铁站火灾人员安全疏散仿真与优化研究 [D]. 辽宁工程技术大学, 2023.

[11]王新宇. 不同起火位置下城市轨道交通地下站点火灾及人员疏散仿真研究 [D]. 重庆: 重庆交通大学, 2024.

[12]焦柳丹, 刘任可, 张羽, 等. 基于可用与必需安全疏散时间的地铁站火灾人员疏散仿真 [J]. 重庆交通大学学报(自然科学版), 2024, 43(10): 106-115.

[13]张蕊. 多场景下地铁站乘客疏散路径规划研究 [D]. 青岛: 青岛理工大学, 2023.

[14]田杰, 苏宇鹏, 吴桐, 等. 基于疏散引导的高峰时段地铁站应急疏散方案优化 [J]. 重庆交通大学学报(自然科学版), 2025, 44(1): 96-103+126.

[15]Xu H, Wei Y, Tan Y. Optimization of emergency evacuation in complex rail transit station [J]. Journal of Building Engineering, 2024, 95: 110321.

[16]Shan S, Guo X, Wei Z, et al. Simulation analysis of evacuation processes in a subway station based on multi-disaster coupling scenarios [J]. International journal of disaster risk reduction, 2023, 96: 103998.

[17]Feng J R, Gai W, Yan Y. Emergency evacuation risk assessment and mitigation strategy for a toxic gas leak in an underground space: The case of a subway station in Guangzhou, China [J]. Safety science, 2021, 134: 105039.

[18]王芳. 基于多智能体的地铁火灾和疏散仿真研究 [D]. 北京: 北京交通大学，2023.

[19]杨晓霞, 蒋海龙, 李永行, 等. 地铁站乘客沿楼梯上行疏散时间预测及安全性评估 [J]. 中国安全科学学报, 2023, 33(5): 168-173.

[20]惠阳, 余强, 王永岗, 等. 地铁站应急疏散效率影响因素分析及优化策略 [J]. 长安大学学报(自然科学版), 2024, 44(4): 130-138.

[21]王勇, 张锦彬, 朱军, 等. 基于改进社会力模型的地铁站火灾应急疏散VR模拟方法 [J]. 中国安全生产科学技术, 2023, 19(5): 194-201.

[22]曲磊. 基于机器学习的地铁站乘客出行时间预测方法研究 [D]. 青岛: 青岛理工大学, 2024.

[23]王杰. 地铁站火灾场景下乘客疏散时间预测与路径优化 [D]. 青岛: 青岛理工大学, 2024.

[24]刘萌. 地铁站火灾情景下人员疏散优化研究 [D]. 西安: 西安理工大学, 2024.

[25]杨大志, 田雨晗. 基于Dijkstra算法的地铁站客流应急疏散路径优化研究 [J]. 今日消防, 2024, 9(8): 4-6.

[26]徐枫. 地铁站安全疏散理论分析及应用 [J]. 中国消防, 2023, (10): 58-61.

[27]Huang L, Li W, Gong J. Simulation of the emergency evacuation about social groups in a complex subway station [J]. Physica A: Statistical Mechanics and its Applications, 2024, 637: 129535.

[28]Cai Z, Zhou R, Cui Y, et al. Influencing factors for exit selection in subway station evacuation [J]. Tunnelling and Underground Space Technology, 2022, 125: 104498.

[29]Zhou R, Cui Y, Wang Y, et al. A modified social force model with different categories of pedestrians for subway station evacuation [J]. Tunnelling and Underground Space Technology, 2021, 110: 103837.

[30]Xu H, Tian C, Li Y. Emergency evacuation simulation and optimization for a complex rail transit station: a perspective of promoting transportation safety [J]. Journal of advanced transportation, 2020, 2020(1): 8791503.

[31]Qin J, Liu C, Huang Q. Simulation on fire emergency evacuation in special subway station based on Pathfinder [J]. Case Studies in Thermal Engineering, 2020, 21: 100677.

[32]张皓. 地铁站内行人异质性应急疏散建模与仿真研究 [D]. 西安: 长安大学, 2024.

[33]从琳. 地铁站应急情况下疏散路径研究 [D]. 石家庄: 石家庄铁道大学, 2023.

[34]袁欣. 基于仿真优化的地铁站行人踩踏风险控制研究 [D]. 武汉: 中南财经政法大学, 2023.

[35]Zhang J, Huang D, You Q, et al. Evaluation of emergency evacuation capacity of urban metro stations based on combined weights and TOPSIS-GRA method in intuitive fuzzy environment [J]. International journal of disaster risk reduction, 2023, 95: 103864.

[36]Tang Y, Xia N, Lu Y, et al. BIM-based safety design for emergency evacuation of metro stations [J]. Automation in Construction, 2021, 123: 103511.

[37]Wang J, Guo J, Wu X M, et al. Study on intelligent algorithm of guide partition for emergency evacuation of a subway station [J]. IET Intelligent Transport Systems, 2020, 14(11): 1440-1446.

[38]陈晨. 基于元胞自动机的低行动力人群疏散仿真研究 [D]. 宁波: 宁波大学, 2021.

[39]任祥霞. 典型场景内老年人多模式疏散特性实验和模型研究 [D]. 合肥: 中国科学技术大学, 2022.

[40]蔡佳良. 多层建筑特殊人群疏散时间及疏散特征研究 [D]. 西安: 长安大学, 2023.

[41]霍非舟, 吕紫含, 李超, 等. 考虑结伴行为的养老场所多群体人员疏散研究 [J]. 灾害学, 2024, 39(3): 153-159.

[42]Murakami H, Feliciani C, Nishinari K. Adaptive Formation by Pedestrian Small Groups During Egresses [J]. Journal of Disaster Research, 2024, 19(2): 379-385.

[43]Takahashi A, Yasufuku K. Evaluation of Tsunami Evacuation Plans for an Underground Mall Using an Agent-Based Model [J]. Journal of Disaster Research, 2024, 19(2): 268-278.

[44]郑杰. 基于元胞自动机的盲人疏散行为建模与仿真研究 [D]. 太原: 太原科技大学, 2022.

[45]王曦晨. 老年人照料设施防火疏散楼梯设计研究 [D]. 大庆: 东北石油大学, 2020.

[46]韩桂荣. 浅谈老年人照料设施的消防疏散 [J]. 中国新技术新产品, 2021, (6): 144-148.

[47]顾正军, 沙鑫烨, 纪勇军, 等. 关于老年人照料设施消防设计相关问题的探讨 [J]. 安装, 2022, (6): 66-69.

[48]曾自立, 陈志艺. 老年公寓建筑火灾危险性辨识及安全疏散分析 [C]//2020中国消防协会科学技术年会论文集. 温州市消防救援支队, 2020: 794-800.

[49]白雪岑. 特殊人群紧急疏散动力学建模与分析 [D]. 太原: 太原科技大学, 2019.

[50]Yazdani M, Haghani M. Elderly people evacuation planning in response to extreme flood events using optimisation-based decision-making systems: A case study in western Sydney, Australia [J]. Knowledge-Based Systems, 2023, 274: 110629.

[51]Sugiyama T, Yamori K. Consideration of evacuation drills utilizing the capabilities of people with special needs [J]. Journal of Disaster Research, 2020, 15(6): 794-801.

[52]V. D S ,G. R S .Peculiarities of calculation of the estimated time of evacuation of people using internal open stairs[J].XXI century. Technosphere Safety,2019,4(2):227-235.

[53]董智龙. 建筑物智能化火灾检测研究 [D]. 大庆: 东北石油大学, 2022.

[54]郭红英. 考虑个体—火场环境因素的养老院人员疏散策略研究 [D]. 赣州: 江西理工大学, 2024.

[55]马辉, 朱红杰, 姚雨婷, 等. 老年人照料设施中失能老人疏散策略研究 [J]. 中国安全生产科学技术, 2020, 16(10): 127-132.

[56]马鑫悦. 养老机构火灾风险评估及保险研究 [D]. 沈阳: 沈阳航空航天大学, 2023.

[57]曹莎莎, 王勤林. 公共交通视角下老年人疏散策略研究—以青岛无棣路历史文化街区为例 [J]. 城市建筑, 2021, 18(33): 185-187.

[58]李昌宇, 曹忠伟, 张若楠. 基于AnyLogic的地铁换乘站客流组织仿真与优化 [J]. 现代城市轨道交通, 2025, (2): 123-129.

[59]房萍萍. 西安大雁塔地铁站集散能力建模仿真分析 [D]. 西安: 西安科技大学, 2020.

[60]全国公共安全基础标准化技术委员会(SAC/TC 351). 地铁安全疏散规范: GB/T 33668-2017 [S]. 中国标准出版社, 2017.

[61]王雷, 霍福民, 王孝坤. 基于Anylogic仿真模型的王府庄站客流组织仿真与优化 [J]. 交通与运输, 2025, 41(1): 20-24.

[62]张思瑶, 杜晨阳, 薛智蒙, 等. 基于Anylogic的地铁站厅客流组织优化研究—以西安地铁小寨站为例 [J]. 河南科技, 2025, 52(2): 62-67.

[63]张龙财, 刘明辉, 刘晨宇, 等. 机场航站楼火灾人员疏散路径研究 [J]. 舰船电子工程, 2025, 45(1): 101-105.

[64]蒋慧灵, 方伟, 徐天锋, 等. 基于Dijkstra算法的室内疏散最优路径规划模型 [J/OL]. 清华大学学报(自然科学版), 1-8[2025-03-18].

[65]刘潇, 王勤勤, 钟导峰. 基于AnyLogic机场安检排队系统仿真优化研究 [J]. 交通科技与管理, 2025, 6(1): 11-13.

[66]刘晓静, 张晨熙. 基于AnyLogic的地铁车站可预见性大客流组织优化方案 [J]. 黑龙江交通科技, 2024, 47(12): 167-170+176.

[67]李慧. 改进蚁群算法的城市地铁应急疏散路径优化 [J]. 中国储运, 2024, (12): 119-120.

[68]牛聚粉, 林雅琳, 楚馨悦, 等. 国内外应急疏散研究进展综述 [J]. 安全与环境学报, 2025, 25(1): 184-195.

[69]Guo C, Huo F, Deng S, et al. An evacuation model considering pedestrian group behavior under violent attacks [J]. Physica A: Statistical Mechanics and its Applications, 2024, 656: 130229.

[70]Li K, Chen Z. Emergency evacuation dynamics based on evolutionary game theory [J]. Physics Letters A, 2024, 528: 130059.

[71]Yang W, Yan X, Hu D, et al. A novel emergency evacuation route optimization model in flood disasters using hydrodynamic model and intelligent algorithm [J]. Safety Science, 2025, 182: 106709.

[72]寻天祥. 地铁换乘站火灾烟气移动与人员疏散研究 [D]. 西安: 西安建筑科技大学, 2024.

[73]张家严. 地铁站火灾事故人员疏散仿真研究 [D]. 长春: 吉林建筑大学, 2024.

[74]倪艳. 基于突发事故的城市地下车站安全疏散能力评价 [D]. 合肥: 安徽建筑大学, 2024.

[75]苗甜甜. 地铁站吊顶对烟气运移的影响及疏散优化研究 [D]. 辽宁工程技术大学, 2022.

[76]朱孔金, 朱军, 王小艳, 等. 考虑换乘通道利用的地铁换乘站行人疏散引导策略研究 [J]. 安全与环境学报, 2024, 24(9): 3548-3555.

[77]吴甜甜, 蒲琪. 地铁车站站台楼扶梯设置参数对行人疏散时间影响 [J]. 城市轨道交通研究, 2024, 27(2): 170-174.

[78]刘意, 王顺利, 胡宏杰, 等. 地铁车站乘客疏散虚拟仿真实验教学设计 [J]. 实验室研究与探索, 2023, 42(12): 79-83.

[79]罗曼, 穆娜娜, 钱小东, 等. 某地铁换乘站火灾人员疏散能力影响分析 [J]. 中国安全生产科学技术, 2023, 19(S1): 177-181.