Figures (7)  Tables (3)

    • Figure 1. 

      Distribution characteristics of hourly passenger flow at different periods.

    • Figure 2. 

      Monthly passenger flow distribution characteristics of rail transit.

    • Figure 3. 

      IPSO-SVR algorithm iterative calculation flow chart.

    • Figure 4. 

      Change curve of model parameter population fitness value.

    • Figure 5. 

      Results of IPSO-SVR passenger flow prediction model. (a) Passenger flow prediction results. (b) Mean square error.

    • Figure 6. 

      MSE curve of the SVR prediction model with the grid method.

    • Figure 7. 

      Results of passenger flow prediction based on different models.

    • Field name Field meaning Data samples
      CARD_ID Card number ***180025
      CARD_TYPE Card type 2
      ENTYR_TIME Entry time 2019***568253
      ENTYR_LINE_NUM Code of entry time 4
      ENTYR_STATION_NUM Code of entry station 5
      EXIT_TIME Exit time 2019***315549
      EXIT_LINE_NUM Code of exit time 4
      EXIT_STATION_NUM Code of exit station 12

      Table 1. 

      Key fields of the Beijing rail transit AFC data sample.

    • Field Field name Data example
      DATE Date 2019/9/11
      PRCP 1-h precipitation (mm) 0
      HUM Relative humidity (%) 56.5
      TEMP Temperature 1.6
      SPD 10-min wind speed (m/s) 4.8
      PRE Atmospheric pressure (HPa) 1014

      Table 2. 

      Main fields of weather data.

    • Prediction models Evaluation metrics
      MSE RA
      IPSO-SVR 1.54% 92.37%
      LSTM 1.71% 89.54%
      SVR 1.94% 87.92%

      Table 3. 

      Results of passenger flow prediction evaluation indexes of different models.