Estimation of Trip-based Generation Models and Calibration of Mode Choice Models for the American Travel Behavior

The National Household Travel Survey (NHTS) is the primary national dataset for analyzing personal and household travel trends. It covers daily non-commercial travel across all modes and details information on travelers, their households, and their vehicles.

The NHTS includes critical data for calibrating travel demand models, such as vehicle details (number and type), individual demographic characteristics (age, gender, employment), and household demographics (income, ownership status, size, race). In this study, we utilized the 2017 NHTS summary statistics to estimate individual trip generation models, differentiating between weekdays and weekends. By comparing factors influencing daily person trips during these periods, we aimed to discern distinctions between the two models. Additionally, we calibrated a mode choice model to understand the impact of trip purpose, duration, household income, and the ratio of available vehicles to household size on the chosen mode. Our analysis focused on identifying and quantifying the factors influencing travel behavior, providing insights into how various variables affect the number of trips and mode choices.

The results indicated variations between weekday and weekend models, with the presence of non-workers and individuals' education levels emerging as crucial factors for weekday travel. Conversely, the existence of children, household income level, and personal yearly miles driven were identified as significant factors affecting weekend travel. Additionally, common characteristics such as household size and urban residence were substantial in both models. The Multinomial regression analysis investigated the correlations between individual, household, activity, and trip characteristics and the modes of transportation selected by travelers. The most significant factors influencing an individual's mode choice are household income, the ratio of available vehicles to household size, and activity purpose.

The study compared weekday and weekend travel behavior using trip-based generation models. Weekday travel was significantly influenced by non-workers and education levels, while weekend travel was more affected by factors like the presence of children, household income, and annual kilometers traveled. Both models emphasized the role of household size and urban residence in shaping travel patterns. The research also examined transportation mode choice, with validation confirming the high accuracy and robustness of the mode choice model in predicting travel behavior.

The study findings are valuable to transportation planners, policymakers, and urban mobility experts aiming to enhance the efficiency and effectiveness of transportation systems. By offering a detailed understanding of individual and household travel patterns, the research enables data-driven interventions that support policy decisions, such as optimizing transit routes, enhancing infrastructure for active transportation, and managing congestion.