Travel Mode Choice Models Based On Bounded Rationality

Urban traffic congestion mainly result from the weekdays' commuting trip,the issue could be relieved by such effective approaches that include adjusting travel mode structure,encouraging residents to take public transportation,as well as limiting the amount of private cars on road.Although the transportation planning and management departments have realized how to develop and regulate,the related policy measures can not be made in a single day.Due to the interrelation of residents' travel choice behaviors,travel environment,road and transportation system,the study of travel mode choice behavior is necessary,which could provide the counting data for transportation demand management.In reality,the decision-making of travelers are bounded rationality,because travel environment usually has non-determinacy and risk,in addition,total traffic information can't be obtained before decision-making.Under the prerequisite of bounded rationality,cumulative prospect theory(CPT)and random utility theory are applied to analyze the mode choice behavior.Different from the discrete choice models,the previous models based on the cumulative prospect theory were indiscrete,the decision makers in which were classified into several categories.In order to break the tradition,the mode choice model based on CPT was developed as discrete form.Firstly,two risk factors were introduced,one is that levy the congestion charges on private cars on some roads,and the other one is that a proportion of travelers by cars would transfer into public transportation,which could increase the crowding in-car.Secondly,the travel mode choice models based on CPT and mixed Logit model were respectively developed,the variables include an indirectly observed factor that is fatigued degree.Assuming that in the risk travel environment,the charged and non-charged roads are available for travelers who choose the car,in addition,they could choose public transport under the uncertainly situation of crowding in-car.Thirdly,the questionnaire was designed to meet the demand of above two models,and the travel data were collected and analyzed.In the empirical research,according to whether have private cars or not,the travelers were divided into two categories.The travel mode choice model based on CPT had been solved to gain each decision-makers' prospect values of car and public transport,and the choice probabilities of the modes in total respondents were got.And then,through the estimation of mixed Logit model to forecast the mode split.The analysis shows that the forecasted probabilities of travelers having private cars on the basis of CPT are approximate to the SP survey result,however the forecasted mode split of mixed Logit model has large departure.It could reveal that when facing the highly punctual commuting mode choice,the application effects of CPT is better than random utility theory,the reason is that random utility theory couldn't distinctly represent the commuters' risk aversion behaviors,that is to say,public transport attract more commuters due to its changeless trip time.