M. Li, Y.H. Wang, L.M. Jia

Pages: 13-22

Abstract
An in-depth research is made on the relevant departments of the typical large urban rail transit, and then the safety elements are analyzed which impact the urban rail transit operation. On the basis of constructing the safety evaluation index system of urban rail transit station operation, in the view of the nonlinear dynamic relationship between safety indexes and evaluation value, the WNN safety assessment model of urban rail transit operation is presented: the AS-OWA aggregation operator method is introduced to dimension reduction so as to simplify the structure of the wavelet neural network; the adjustment algorithms based on hidden layer nodes is used to determine the number of wavelet basis in order to reduce the wavelet neurons of the redundancy of information and improve the convergence speed. Taking Beijing rail transit typical station operation as the object, based on MATLAB platform, the WNN safety assessment simulation model of urban rail transit operation are built to assess the station operation. The results show that the WNN is an effective method in the training accuracy and iterations.

Keywords: urban rail transit; station operation; safety assessment; index system; WNN