地铁站
行人
紧急疏散
运输工程
楼梯
地铁站
计算机科学
工程类
土木工程
海洋学
地质学
作者
Ying Lu,Yunxuan Deng,Shuqi Sun
出处
期刊:Engineering, Construction and Architectural Management
[Emerald (MCB UP)]
日期:2023-07-25
卷期号:31 (9): 3479-3507
被引量:4
标识
DOI:10.1108/ecam-12-2022-1169
摘要
Purpose Metro stations have become a crucial aspect of urban rail transportation, integrating facilities, equipment and pedestrians. Impractical physical layout designs and pedestrian psychology impact the effectiveness of an evacuation during a metro fire. Prior research on emergency evacuation has overlooked the complexity of metro stations and failed to adequately consider the physical heterogeneity of stations and pedestrian psychology. Therefore, this study aims to develop a comprehensive evacuation optimization strategy for metro stations by applying the concept of design for safety (DFS) to an emergency evacuation. This approach offers novel insights into the management of complex systems in metro stations during emergencies. Design/methodology/approach Physical and social factors affecting evacuations are identified. Moreover, the social force model (SFM) is modified by combining the fire dynamics model (FDM) and considering pedestrians' impatience and panic psychology. Based on the Nanjing South Metro Station, a multiagent-based simulation (MABS) model is developed. Finally, based on DFS, optimization strategies for metro stations are suggested. Findings The most effective evacuation occurs when the width of the stairs is 3 meters and the transfer corridor is 14 meters. Additionally, a luggage disposal area should be set up. The exit strategy of the fewest evacuees is better than the nearest-exit strategy, and the staff in the metro station should guide pedestrians correctly. Originality/value Previous studies rarely consider metro stations as sociotechnical systems or apply DFS to proactively reduce evacuation risks. This study provides a new perspective on the evacuation framework of metro stations, which can guide the designers and managers of metro stations.
科研通智能强力驱动
Strongly Powered by AbleSci AI