The accuracy of short-term passenger flow prediction for urban rail transit is of great significance for the control of passenger flow at urban rail stations during the morning and evening peaks on weekdays,as well as for the decision-making of passenger travel programs.Therefore,a short-time passenger flow prediction model of urban rail transit using convolutional neural network(CNN)combined with bidirectional long-short-term memory neural net-work(Bi-LSTM)is proposed based on deep learning,which combines the bidirectional extraction capability of Bi-LSTM for time series data features,and at the same time,CNN is employed in realizing the local feature extraction of data,so as to enhance the generalization capability of the model.The weekday passenger flow data from Xujiahui Sta-tion of Shanghai Metro Line 1 is selected as the basis for optimization of the number of layers and convolution kernel length of the CNN layers,and the ablation and comparison experiments are conducted for the model,respectively.The experimental results show that the prediction accuracy of CNN-Bi-LSTM is optimal,and the RMSE and MAPE are 16.699 9 and 13.52％,respectively,which proves the effectiveness of the model in predicting the short-term pas-senger flow of urban railways.
维普
万方数据