Simulation Study On Fire Smoke Migration And Personnel Evacuation In Island Subway Station Based On Scenario Evolution

In recent years,the construction of municipal railroad transportation has developed rapidly in our country.Owing to the advantages of large passenger flow and punctuality,it greatly facilitates people’s daily travel.However,as the interchange of railway transit lines,subway stations not only gather many passengers,but also located in personnel-intensive sites.There are many combustible factors in the subway system and the space is relatively closed.Once the fire accident occurs,it will certainly cause extremely serious consequences.In addition,subway station fire accident,as an emergency,has the characteristics of bruptness,uncertainty and dynamics,so it is essential to take scientific and effective emergency measures to deal with it.Therefore,it is particularly important how to predict the evolution path and development trend of the accident reasonably and effectively according to the current state,and then adopt scientific and effective emergency disposal methods to deal withIn this paper,based on the key scenario states in the process of accident evolution and their corresponding disaster-pregnant environment,emergency activities,disposal objectives and other elements,the dynamic Bayesian network method was constructed.Selecting the factor nodes with greater influence were selected by the sensitivity analysis of the key scenarios accident state.Based on scenario construction and deduction,PyroSim and Anylogic numerical simulation software are employed to the three-dimensional reconstruction of fire parameters and passenger evacuation,aims to provide technical support for the scientific prevention and emergency treatment of subway fire accidents.The specific research is as follows:（1）Based on the analysis of typical subway fire disaster mechanism,the consideration of actual structure of station and knowledge element theory,the elements of subway station fire accident scenario are extracted.The traditional Bayes is extended to the fuzzy field to overcome the limitation and subjectivity of probability allocation,and to meet the characteristics of ambiguity,uncertainty and information incomplete in the development of subway station fire accidents.Node probability calculation and sensitivity analysis are carried out by GeNIe 2.0 software,and importance analysis is carried out from the level of situational elements.（2）PyroSim numerical simulation software is used to analysis of the interlacing of multiple scenarios with disast-pregnant environment E₇"smoke filled,high temperature and instability of support body"and emergency activity A₇"smoke exhaust system opening and water curtain protection".Through the reconstruction of the fire process,the numerical data of fire products at different times of the platform and the station hall floor in the simulation period of 600s are obtained,and the distribution law of fire products was analyzed.At the same time,the influence of water mist and smoke exhaust system on the distribution characteristics of CO concentration and temperature field are explored,and suggestions for improvement of emergency activity A are put forward.（3）Select the disaster pregnant environment E₇"smoke filled,high temperature and instability of support body"and S₈"personnel evacuation"key accident scenario,based on the numerical data of various fire products obtained from smoke simulation of subway station fire,and comprehensively consider the influence of instability of support body in the fire process.Through Java programming and Anylogic software,the setting of passenger evacuation speed at different time and position within 360s under fire scenario E₇ is realized.The evolution process of S₈"personnel evacuation"under fire scenario is simulated within 360s,and corresponding improvement suggestions are put forward for the problems existing in the process of passenger evacuation.