Identifying and Analyzing Highway Commuter Vehicles Based on Regular Travel Behaviors

With the continuous expansion of urban agglomerations, more and more travelers are relying on highways for commuting. Identifying highway commuter vehicles and analyzing their travel behaviors can help alleviate highway congestion and provide valuable insights for policy-making. Existing identification methods primarily focus on urban roads and often rely on rule design or simple feature extraction, which may not fully capture the complexity of highway commuting behaviors. In this paper, we propose a novel highway commuter vehicle identification method based on regular travel behaviors. This method first merges adjacent travel regions for individual vehicles, and then employs a spatiotemporal clustering algorithm to extract trips with similar spatiotemporal characteristics as regular travel behaviors. Seven features are extracted from these regular behaviors and used as input to the K-means algorithm to identify highway commuter vehicles. We conducted extensive experiments using a three-week highway trip dataset from Chongqing, China. The analysis revealed two distinct types of commuter vehicles: Type I, which typically commutes 2-3 days per week over long distances between different districts, exhibiting a weekly commuting pattern; and Type II, which commutes about 5 days per week over shorter distances within or between adjacent districts, exhibiting a daily commuting pattern.