Research On Railway Emergency Decision Methods Based On Design Structure Matrix And Bayesian Decision Network

As an indispensable modern means of transportation,railway is an important framework of modern urban operation,which is of great significance to China's economic development,national travel,freight transportation and other aspects.With the rapid development of China's railway,it is undertaking more and more important task of large passenger transport.At the same time,the demand of passengers for rapid handling of emergencies is also increasing.Due to the characteristics of railway emergencies,its emergency decision-making mode is generally based on the contingency plan,but the deviation between the actual situation and the scenario of the contingency plan is inevitable,so the rescue usually depends on the guidance of the emergency rescue command group,and the initial decision-making may not be able to control the development of the incident perfectly,which requires the decision makers to change the plan flexibly to achieve the best emergency effect.This thesis takes the emergency decision-making method of railway emergency as the research object,and according to the general disposal way of railway emergency,obtains the emergency disposal process under the specific situation.On the basis of the process,this thesis analyzes the dependence relationship between the emergency activities of the process,constructs the design structure matrix model for the dependence relationship,transforms it into the emergency decision network of the emergency event,analyzes the alternative schemes of the nodes by the utility risk entropy,and provides the basis for the dynamic adjustment of the scheme.First of all,from the whole process of high-speed railway emergency response,combined with a specific emergency plan and emergency drill process,this thesis summarizes the emer-gency response process.The emergency disposal part is analyzed as an example,which is transformed into UML activity diagram.According to the mapping between activity diagram and DSM,the dependence relationship between emergency activities is obtained and ex-pressed in DSM model.Secondly,by introducing the idea of agent,agent is used to replace decision-making sub-ject,and Bayesian decision network is used to reflect the relationship among decision-making subject,decision-making and decision-making scheme in the form of network.According to the dependency expressed by the design structure matrix model,combined with the Bayesian decision network,the corresponding railway emergency decision network is established to express the decision and its relationship that can be made in response to the emergency.Thirdly,considering the complexity and variability of railway emergencies,the emergency process of railway emergencies is divided into different stages.At each stage,the emergency effect of the current scheme is analyzed,and the risk entropy is obtained when the psychological expectation of the decision-maker and various possible outcome scenarios are considered in the emergency scheme set,which provides guidance for the decision-maker to adjust the scheme dynamically in different stages.Finally,this thesis designs and implements the railway emergency decision supportsystem,designs the overall operation framework and decision-making process of the system,shows the function of the system,and analyzes the operation results to verify the feasibility of the method and the effectiveness of the system.