The conventional traffic demand forecasting methods based on revealed preference (RP) data are not able to predict the modal split. Passengers' stated intentions are indispensable for modal split forecasting and ev...The conventional traffic demand forecasting methods based on revealed preference (RP) data are not able to predict the modal split. Passengers' stated intentions are indispensable for modal split forecasting and evaluation of new traffic modes. This paper analyzed the biases and errors included in stated preference data, put forward the new stochastic utility functions, and proposed an unbiased disaggregate model and its approximate model based on the combination of RP and stated preference (SP) data, with analysis of the parameter estimation algorithm. The model was also used to forecast rail transit passenger volumes to the Beijing Capital International Airport and the shift ratios from current traffic modes to rail transit. Experimental results show that the model can greatly increase forecasting accuracy of the modal split ratio of current traffic modes and can accurately forecast the shift ratios from current modes to the new mode.展开更多
Traditional trip generation forecasting methods use unified average trip generation rates to determine trip generation volumes in various traffic zones without considering the individual characteristics of each traffi...Traditional trip generation forecasting methods use unified average trip generation rates to determine trip generation volumes in various traffic zones without considering the individual characteristics of each traffic zone. Therefore, the results can have significant errors. To reduce the forecasting error produced by uniform trip generation rates for different traffic zones, the behavior of each traveler was studied instead of the characteristics of the traffic zone. This paper gives a method for calculating the trip efficiency and the effect of traffic zones combined with a destination selection model based on disaggregate theory for trip generation. Beijing data is used with the trip generation method to predict trip volumes. The results show that the disaggregate model in this paper is more accurate than the traditional method. An analysis of the factors influencing traveler behavior and destination selection shows that the attractiveness of the traffic zone strongly affects the trip generation volume.展开更多
文摘The conventional traffic demand forecasting methods based on revealed preference (RP) data are not able to predict the modal split. Passengers' stated intentions are indispensable for modal split forecasting and evaluation of new traffic modes. This paper analyzed the biases and errors included in stated preference data, put forward the new stochastic utility functions, and proposed an unbiased disaggregate model and its approximate model based on the combination of RP and stated preference (SP) data, with analysis of the parameter estimation algorithm. The model was also used to forecast rail transit passenger volumes to the Beijing Capital International Airport and the shift ratios from current traffic modes to rail transit. Experimental results show that the model can greatly increase forecasting accuracy of the modal split ratio of current traffic modes and can accurately forecast the shift ratios from current modes to the new mode.
基金the National Natural Science Foundation of China (No. 50478041)the Natural Science Foundation of Beijing (No. 8053019)
文摘Traditional trip generation forecasting methods use unified average trip generation rates to determine trip generation volumes in various traffic zones without considering the individual characteristics of each traffic zone. Therefore, the results can have significant errors. To reduce the forecasting error produced by uniform trip generation rates for different traffic zones, the behavior of each traveler was studied instead of the characteristics of the traffic zone. This paper gives a method for calculating the trip efficiency and the effect of traffic zones combined with a destination selection model based on disaggregate theory for trip generation. Beijing data is used with the trip generation method to predict trip volumes. The results show that the disaggregate model in this paper is more accurate than the traditional method. An analysis of the factors influencing traveler behavior and destination selection shows that the attractiveness of the traffic zone strongly affects the trip generation volume.