Risk Assessment of Subway Station Fire By Using a Bayesian Network-Based Scenario Evolution Model
Autor(en): |
Xuewei Li
Jingfeng Yuan Limao Zhang Dujuan Yang |
---|---|
Medium: | Fachartikel |
Sprache(n): | Englisch |
Veröffentlicht in: | Journal of Civil Engineering and Management, 12 März 2024, n. 3, v. 30 |
Seite(n): | 279-294 |
DOI: | 10.3846/jcem.2024.20846 |
Abstrakt: |
Subway station fires frequently result in massive casualties, economic losses and even social panic due to the massive passenger flow, semiconfined space and limited conditions for escape and smoke emissions. The combination of different states of fire hazard factors increases the uncertainty and complexity of the evolution path of subway station fires and causes difficulty in assessing fire risk. Traditional methods cannot describe the development process of subway station fires, and thus, cannot assess fire risk under different fire scenarios. To realise scenario-based fire risk assessment, the elements that correspond to each scenario state during fire development in subway stations are identified in this study to explore the intrinsic driving force of fire evolution. Accordingly, a fire scenario evolution model of subway stations is constructed. Then, a Bayesian network is adopted to construct a scenario evolution probability calculation model for calculating the occurrence probability of each scenario state during subway station fire development and identifying critical scenario elements that promote fire evolution. Xi’an subway station system is used as a case to illustrate the application of Bayesian network-based scenario evolution model, providing a practical management tool for fire safety managers. The method adopted in this study enables managers to predict fire risk in each scenario and understand the evolution path of subway station fire, supporting the establishment of fire response strategies based on “scenario–response” planning. |
- Über diese
Datenseite - Reference-ID
10774195 - Veröffentlicht am:
29.04.2024 - Geändert am:
29.04.2024