Research On The Model Of Emergency Evacuation Capability Evaluation For Subway Station Based On Bayesian Network

With the rapid development of urbanization,the urban ground traffic has undergone great pressure.The Subway,due to its fast speed,high effectiveness and no occupation of ground traffic resource,is highly appreciated by the government,expert and the public.However the complex and equipment-crowd environment,besides the confined space and large passenger flow,would make it vulnerable to those emergency accidents,such as fire,explosion,power outage,terrorist attack,and subway collision.Once we fail to evacuate in time,it will cause malignant chain reaction,including smoke poisoning,stampede,over crowding etc,which will greatly threaten public life safety together with national wealth.Therefore,the scientific way to recognize factors that affects subway evacuation and make a judgment about evacuation ability can not only form a clear understanding about it but also help to improve the following evacuation performance.This study,with the analysis on subway evacuation accident and relative literature reading,making a summary about various factors that affect subway emergency evacuation according to accident causation theory,sifting those factors based on the principle of index establishment to establish the subway emergency evacuation ability assessment system.Secondly,the evaluation index system is transformed into a Bayesian network topological structure according to the causal relationship.Combing with the characteristics of Bayes network and subway station emergency evacuation ability,we propose weight-based method of calculation for conditional probability,get the child node conditional probability table.Give the root node edge probability acquisition and calculation method are obtained.And we can explain the method and function of the model reasoning.At last,taking Zhengzhou Huanghe road station as an example,the paper evaluates the emergency evacuation capability of the station with the established Bayesian network model.By using diagnostic reasoning,the weak links in the evacuation of the station are found out,targeted countermeasures and suggestions are put forward.the fuzzy comprehensive evaluation of the station’s emergency evacuation capability is carried out,the evaluation results are compared with the evaluation results of the Bayesian network model,the correctness and effectiveness of the Bayesian network model is verified,and its advantages are analyzed.Based on the safety and risk assessment of the existing Bayesian Network,this paper aims at proposing weight-based method of calculation for conditional probability.We hope a better application of Bayes network on evacuation ability assessment.