WIT Press


Generation Of Emergency Scheme For Urban Rail Transit By Case-based Reasoning

Price

Free (open access)

Volume

114

Pages

8

Page Range

529 - 536

Published

2010

Size

379 kb

Paper DOI

10.2495/CR100491

Copyright

WIT Press

Author(s)

F. Li, R. Xu & W. Zhu

Abstract

Case-based reasoning (CBR) is a method that uses previous experiences to solve new problems. The characteristics of CBR make it suitable for complex problems related to knowledge reuse. Based on the analysis of the characteristics of emergency events in urban rail transit, a generation method using CBR for emergency scheme is proposed. Certain related technologies, such as case representation, case retrieval, case adaption, case revising and case retaining, are conceived and discussed. A numerical example is used to illustrate the application and efficiency of the proposed method, which can take full advantage of historical correct experiences and benefit to the intelligentization of emergency scheme generation for urban rail transit. Keywords: urban rail transit, emergency scheme, case-based reasoning, case retrieval. 1 Introduction Accidents, failures, unpredictable disasters, and sudden increase of passenger flow due to special events happen in urban rail transit (URT) now and then. These emergency events, which impact widely and long and evolve uncertainly, will bring negative impact on the whole rail network unless properly dealt with. In terms of URT, on one hand, operation organization and passenger evacuation are constrained by the spatial layout because of the relatively closed space in URT system. On the other hand, emergency response process for URT actually is a process of interaction and cooperation between different departments, which is difficult to do well due to the distinct differences existing in the communication, coordination and authorization. Therefore, it is visible that the emergency decision in URT is complex and difficult.

Keywords

urban rail transit, emergency scheme, case-based reasoning, case retrieval