In view of the fact that most of the current short-term passenger flow forecasts rely only on inbound passenger flow, a sudden arrival passenger flow prediction model is proposed with two factors, the outbound and inbound passenger flow caused by activities. The wavelet neural network (WNN) model is selected and optimized by the genetic algorithm (GA), according to two-factor data characteristics. Taking the Beijing Dongsishitiao Subway Station as an example, the sudden passenger flow events from 2014 to 2016 are taken as samples to train and verify the reliability of the prediction model. The result shows that the optimized WNN is better than the traditional WNN, and the error of models based on two-factor is significantly smaller than models with single-factor. The proposed model has high accuracy, which can meet the needs of sudden passenger flow management of stations with large venues nearby.

CICTP

暂无