Research On Railway Emergency Decision Making Based On CBR

In recent years,with the rapid development of China railway,railway emergency has also occurred sometimes,which poses a serious threat to people's lives,property,and social stability.For the emergency handing of railway emergency,the current decision-makers mainly solve the current problems based on the experience of dealing with past incidents.However,due to the knowledge limitations of the decision-makers and their own psychological pressure,it may cause an effective and reasonable solution not to be achieved in a short time.Therefore,how to make effective decisions on railway emergency efficiently and timely has become an urgent problem to be solved.Case-Based Reasoning(CBR)is an analogical reasoning method that solves current problems based on past practical experience.The process of solving the problem is consistent with the basic ideas of humans to solve the problem.It can quickly retrieve the similar cases from the past,obtain reference solutions more accurately,which fully utilize the experience of past cases to solve the current problems.Therefore,this paper introduces CBR into railway emergency decision-making,and proposes a method of emergency decision-making for railway emergency based on CBR,and studies its key technologies.The main research content of this paper is as follows:(1)The similarity between CBR and railway emergency in the problem-solving process has been studied.Firstly,through literature analysis,the connotation,classification,and characteristics of railway emergency were extracted;then the key technologies and characteristics of CBR were studied,and the definition and process of railway emergency decision-making were given.Finally,based on this analysis,the similar relationship between CBR and railway emergency decision-making process is analyzed.(2)The case attributes and representation methods of railway emergency cases were studied.Firstly,through literature analysis,has obtained the subject domain,problem domain,solution domain,and the effect domain.Then compare the advantages and disadvantages of the currently used case representation methods,and then use the framework method to express the railway emergency cases.(3)Researched the retrieval and the generation method of emergency plans.Firstly,in order to improve the retrieval efficiency and make the retrieval result accord with the actual situation,this paper adopts the dual retrieval mechanism which contains hierarchical retrieval and nearest neighbor retrieval.When performing the nearest neighbor search,the core lies in the similarity calculation.Considering the structural characteristics of the railway emergency,a Overall case similarity calculation method which has double-layer structure based on structural similarity and attribute similarity is introduced based on the traditional nearest neighbor search algorithm.Taking into account the different types of attributes,the similarity calculation methods for symbolic attributes,numeric attributes,and fuzzy variable attributes are given.The attribute similarity calculation needs to consider the attribute weight.This paper analyzes the advantages and disadvantages of the commonly used weighting methods and combines the expert scoring with the cloud model to get attribute weight.Finally,the generation method and the steps of the emergency plan were analyzed,and verified by example analysis.(4)Based on CBR,the railway emergency decision system was designed and developed.In order to improve the efficiency and scientificity of railway emergency decision-making,the functional modules such as case management,case retrieval,program evaluation,program revision,system help and maintenance were designed,and a demonstration of the system's functions was performed through examples.