Understanding the route choice behavior of truck drivers in urban transportation delivery networks is essential for enhancing freight mobility and optimizing route planning strategies. This study introduces a data-driven method that integrates route choice set generation with discrete choice modeling (DCM) to identify the key factors influencing truck drivers' route selection. Based on the truck GPS trajectory data in the Amsterdam metropolitan area, route choice sets are generated using Monte Carlo Labeling (MCL) and Accelerated Monte Carlo (AMC) methods. By using the generated choice sets as inputs and considering the influencing factors of road grade, route distance, number of turns, and route size, a Path Size Logit (PSL) model is established to estimate the impacts of different factors. The findings contribute to a deeper understanding of truck route choice behavior and offer practical insights for the development of smart route recommendation systems for logistics operators.
