Emergency System For Urban Rail Transit Based On Case-Based Reasoning

The emergencies have the characteristics of sudden, urgency and uncertainty. Therefore, when the emergency happens, the decision maker is required to provide accurate and effective emergency decision under the condition of limited time and information. In addition, the result of the emergency strategy is difficult to predict, which will impose the decision maker a huge psychological pressure. In general, reliable emergency aid decision-making system has important application value in providing decision support.The research methods and theories on the problems of train operation adjustment under the condition of urban rail transit operation emergency were systematically discussed and analyzed. The main researching contents are:Firstly, this paper reviews the research results on urban rail traffic emergencies and reasoning methods by cases at home and abroad and concludes the ways, strategies and the relative key techniques in reasoning by cases which are used in train operation adjustment under the condition of urban rail transit operation emergency.After making a statistical analysis on main urban rail traffic emergency at home and abroad, a conclusion is made that urban rail traffic emergencies at home and abroad has different type distribution and an explanation is made for this phenomenon. According to the characteristics of the domestic main urban rail traffic emergencies, the emergencies can be divided into four scenes. Based on the researches on urban rail transit emergency event, such as case representation, case retrieval, case revision, case reservation and case maintenance, discusses the adjustment and modification based on the frame representation of city rail traffic emergency event case represents the process of emergency plan.This paper adopts a case similarity calculation method based on distance membership function to calculate and analyze the similarity of case attributes, the local similarity of case attributes. AHP, entropy method, gray correlation theory and the K-Nearest Neighbor algorithm are applied to calculate the case attribute weights and the global similarity of emergency cases.Decision model and design on emergency system is proposed by creating a database of urban rail traffic emergency events and combining the experience and knowledge in this field. The overall design scheme of urban rail transit emergency system based on case-based reasoning is proposed, and the design of system function module, the work flow of the system are mainly introduced and the development of emergency system based on PHP and MySQL is discussed in this paper.Finally, this paper creates a database on the basis of historical data of urban rail transit operation. The decision model of this paper and emergency system are realized by the field experience knowledge.All of this provide important theoretical bases and practice instruction for emergency decision making in urban rail transit.