Mode Choice Modeling Using Adaptive Data Collection for Different Trip Purposes in Mumbai Metropolitan Region

摘要

The issues of traffic and transportation planning in cities such as Mumbai are wellknown, but they are not well understood.Understanding mode choice behavior of travelers is critical for transportation planners. This paper focuses on understandingthis behavior and understanding the factors that influences mode choice in Mumbai Metropolitan Region. An online surveysystem was designed to enable the users to fill in the revealed choices they made for a particular trip. This research examinesthe impact of socio-demographic attributes of travelers along with trip characteristics, such as travel time and waiting timeon mode choice behavior. The data include 253 trips undertaken by 178 persons for three major trip purposes: mandatory,shopping, and recreational was captured. Multinomial logit (MNL) and mixed logit (MXL) models are developed to portraymode choice behavior. The models are able to capture the transit-captive behavior of commuters. The MXL model slightlyoutperformed the MNL model both, in terms of statistical fit and prediction success rate. This research exemplifies that choiceof mode is often governed by purpose associated with the activity. Considering that there has been a huge modal shift frompublic transit to private vehicles (cars and two-wheelers) in Mumbai over the last decade, estimation of mode choice modelsare important for policy formulation.