Transfer Penalty of Public Transport

The public transport accounts for a large proportion in urban traffic, however, generally, travelers are unable to complete the trip with a single form of public transport. With the rapid development of the intelligent transportation and the application of big data, the accuracy of the generalized traffic time is more and more important to the passenger flow forecasting and service level optimization of public transportation system. An improved Logit model based on the theory of principal component analysis is established to calibrate the transfer penalty, which can cover the correlation among the variables. The calibration results show that the improved model is more accurate than the traditional Logit model, the conclusion is more reliable. In this paper, when the total time of the options is similar, if taking the different number of bus line into consideration, the transfer penalty is 12.13 minutes when the number of bus line is one, while it's 9.93 minutes when the number of bus line is two; if taking the same number of bus line into consideration, when there is only one bus line, the transfer penalty is 10.13 minutes, while it's 7.92 minutes when the number of bus line is two.