A Situation Assessment Model For Decision-making And Supervision Of Railway Emergency Response

Railway transportation,serving as the framework of popular means of transportation in modern city,has been of great significance for the development of Chinese economy,homeland security,life quality as well as the comprehensive transportation system,where safe transportation tends to play the key role as Chinese railway transportation.By far,the Ministry of Railways of China has established a railway emergency management system attached importance to planning,regulation,institution and legislation for emergency rescues(known in Chinese as ―One Planning plus Three Systems‖).Rescuing plans should be made for different cases after a sound plan was made based on the previous experience.However,as the technology,interaction between systems and the complexity escalates,the task for rescuing has become more urgent and difficult.As a result,the efficiency to handle emergencies had become the key factor to evaluate the railway’s ability to come up with contingency plans.Rescue plans should be the contingency plans that had considered all possible scenarios and the plans that involve the cooperation of all the departments.Based on the aforementioned background,the contingency plans that is conceived based on the recovery of the real case scenario can effectively serve the railway system and make a profound impact on improving the management skills for the railway industry.This essay will discuss the contingency plans and railway supervision problems in response to the emergencies on railway based on the study of the findings and the current conditions of the emergency responses from real scenarios from China and other countries,historical case studies and the method of probability theory and informational theory.Firstly,this essay evaluates the terms and procedures for emergency responses in current emergency management system.By comprehensively analyzing the cases for emergencies,this essay will sort these accidents based on the characteristics of the cases and the quantitative and qualitative cloud model evaluation.Secondly,this article will analyze the probability for variation for these cases using a multi-factor evaluation method.By analyzing the effectiveness in response to the emergency cases and the state variation in real scenarios,a dynamic vectoral Bayesian network mapping from probability graph model was applied to evaluate this state variation using a dynamic state model to predict its tendency in response to the escalation for the emergency and providing support for the contingency plans.In case of a derailed train,a dynamic vectoral Bayesian network mapping can be applied to analyze the track damage level.The result from the tendency analyzing result matched perfectly with the real case scenario.Then contingency plans in response to the tendency analyzing results were studied,which is achieved by the following procedures: improving the distance entropy weight method and therefore assess the weight corresponding to the properties to the case thus optimizing the attributes to the problem.Using single value decomposition(SVD)to lower the dimension of the case information to reduce the complexity of the calculation.Using clustering algorithm to cluster the case information to improve the calculation.Finally,by searching similar cases,additional information of responding level,rescue plans,organizing method as well as emergency resources were gathered from these cases.The analyses show that this algorithm is convenient and highly achievable in solving these cases.Also,since this model will process the case statistics before the emergency occurs,therefore,the complexity and the running time is greatly reduced,which agrees with the requirement of solving urgent emergency cases.Finally,in response to the current emergency response system which relies greatly on the human experience and subjective idea,as well as noticing the fact that some problems may not be spotted while emergency departments were following the inflexible procedures.This article also constructs an emergency monitoring and predicting model based on random network Graphical Evaluation and Review Technique(GERT),which could provide straight forward criteria for rescuing techniques,rescuing procedures and systematic evaluation for the contingency plans by also considering the background information,staff members as well as management issues for the emergencies.