Abstract:
Mode choice or modal split is the fourth step of transportation planning. The step discuss with
the mode choice behavior of road user for a particular set of origin and destination node specified
to particular route. Modal split depends on travel cost, travel time, convenience and comfort.
There are some other factors which may influence mode choice behavior of road users. Mode
choice for long distance travel is different than daily short distance trip. Long distance trips are
conducted by the road users for once or twice a week or month and it is not frequent. On
contrary, short distance trips are regular and road user do the trips once or more than twice
sometimes in a day. Few previous studies have been performed on long distance modal split
behavior of trips. No prior studies have been conducted in context of Bangladesh. Our study is
unique in this point of view. We have considered four routes around Dhaka city. We have done
questionnaire survey for Dhaka to Chittagong route, Dhaka to Sylhet route, Dhaka to Khulna
route and Dhaka to Rajshahi route. We have done stated preference survey on 302 respondents.
The respondents are of several age group, income level, occupation and gender group. The
independent variables of our study are, Travel time, Travel cost, Comfort, Safety, Reliable
Journey time, Stop/station closer from origin, Stop/station closer to destination, Availability of
tickets and Flexibility of schedule. The dependent variable is mode choice preference. In this
study our alternative modes are intercity bus and intercity train. We developed artificial neural
network (ANN) model to rank the importance of the variables. The variables ranking and its
importance will be used by the city transportation planner for policy making.