Innovative Data Collection and Modeling Methods for Long-Distance Passenger Travel Demand Analysis

After the Intermodal Surface Transportation Efficiency Act was established in 1991, an increasing number of state highway agencies and federal agencies have started to develop and implement statewide or national travel demand models to meet policy and legislative development needs, and to predict the future travel demand. To date, more than 35 states have conducted modeling developments at the statewide level (Cohen, Horowitz, & Pendyala, 2008; Giaimo & Schiffer, 2005; Horowitz, 2006, 2008; Souleyrette, Hans, & Pathak, 1996). However, a lack of up-to-date multimodal and inter-regional travel survey data hinders researchers’ or analysts’ ability to quantitatively conduct reliable and effective evaluation of long-distance travel infrastructure investment and management at the statewide level. Meanwhile, in Europe travel demand modeling at the national level has received more attention in the last two decades. From the perspective of geography and population size, the European national travel demand model, to an extent, can be taken to be a statewide model in the U.S. Among the efforts involved in long-distance passenger travel modeling, the travel data collection is found to play a critical role in the success of the travel demand modeling at both the statewide and national levels. In this report, the post-processing methods (machine learning methods) to automate the trip purpose estimation are developed for long-distance travel, and available datasets including travel survey data and other supplementary data are employed to test and validate the method. This research aims to provide the support tool for long-distance travel data collection and sound methodology for post-processing the Global Positioning System (GPS)-, smartphone-, and social media-based travel survey data in the future. Alternative trip purpose categorization schemes for long-distance travel have been developed. Furthermore, the model performance under different purpose categorization is tested in order to provide comprehensive information to assist the design of future long-distance travel surveys.