Prediction Model of Passenger Tolerable Waiting Time Under Metro Train Delay Condition

Abstract:

To deepen researches on metro passenger travel behavior under abnormal operating conditions， a prediction model of different passengers' tolerable waiting time under the metro train delay condition based on the survival analysis theory was proposed. First， the passenger tolerable waiting time was defined， and its characteristics were analyzed. Then， a parameter model based on Weibull distribution was selected to build the passenger tolerable waiting time prediction model considering various influencing factors. Finally， the proposed model was calibrated by the data obtained from a survey in combination with revealed preference and stated preference. The results show that the willingness of passengers to give up waiting and choose to leave the station increases along with the waiting time. Moreover， the passenger tolerable waiting time is positively correlated with the use frequency of metro per week， the metro travel time of the daily trip， the buffer time of the whole trip， and the metro travel time of the current trip. Compared with traveling for entertainment or other purposes， passengers have less tolerable waiting time when commuting. The probability that the tolerable waiting time of metro-oriented commuters exceeds 5.4 min and 14.3 min is over 80 % and below 50 %， respectively.

Key words:urban rail transit;train delay;passenger tolerable waiting time;survival analysis;Weibull distribution