Mode Choice Indicators Study Of P&R Using The Latent Class Model

Studying the indicators of mode choice as well as analyzing the external and latent factors of different travel classes are important for the stockholders,it could help them making more feasible transport policies,release the traffic pressure,rational allocation of road resources and improving urban life quality.Park and ride system is an important traffic demand management.The system has been used and studied worldwide.However,the usage percentage of each P&R is vary due their own situation.There are a lot of factors affecting Park and Ride system,include the park and ride setting and traveler characters.The study is based on the Park and Ride station propose in Beverly,UK.The writer use the binary Logit model to assess the external indicators of Park and Ride system.And then,based on the results of the binary Logit modal,a stated preference questionnaire was been held to collect the data.The data is used to carry on a Latent class model.In the first place,the research used the Binary Logit Model to analysis the elastic of five different factors,increasing the parking fee(city center),reducing the parking space,improving the bus journey time,reducing the bus fee and the penalty value.From the initial assessment,the most effective way of improving the proportion of park and ride system is increasing the out-of pocket cost by car.And then follow by bus journey time and parking space reduction.The factors elastic is used to design the Hypothetical scenario of the questionnaire.The data collected from the questionnaire is people`s travel behavior.Finally,the Latent class model is able to provide 4 latent classes and the probability of each class.The model results shows the potential users of Park and Ride system have high reaction rate to the comfortable and convenience of P&R station.