Analysis of Key Commuting Routes Based on Spatiotemporal Trip Chain

Abstract

Commuting pattern is one of the most important travel patterns on the road network; the analysis of commuting pattern can provide support for public transport system optimization, policy formulation, and urban planning. In this study, a framework of the key commuting route mining algorithm based on license plate recognition (LPR) data is proposed. And the proposed algorithm framework can be migrated to any similar spatiotemporal data, such as GPS trajectory data. Commuting pattern vehicles are first extracted, and then, the spatiotemporal trip chains of all commuting pattern vehicles are mined. Based on the spatiotemporal trip chains, the spatiotemporal similarity matrix is constructed by dynamic time warping (DTW) algorithm. Finally, the characteristics of commuting pattern are analysed by the density-based spatial clustering of applications with noise (DBSCAN) algorithm. Different from other researches that analyse the commuting pattern using machine learning algorithms based on all data, this study first extracts commuting pattern vehicles and then designs a key commuting route mining algorithm framework for commuting pattern vehicles. Taking Hangzhou as an example, through the framework of mining algorithm proposed in this study, the commuting pattern characteristics and key commuting routes in Hangzhou have been successfully excavated, and policy suggestions based on the analysis results have also been put forward.

Understanding travel behavior adjustment under COVID-19

“…This study uses POI data to reflect the type of land usage around a certain LPR detector. The method proposed by Yao et al (2022) is used to analyze the land use types, and the method is improved combined with the method proposed by Gao et al (2021b) . The method mainly includes the following three steps: The category of POI is reclassified to better reflect the land use type.…”

Clustering Study of Vehicle Behaviors Using License Plate Recognition

“…In [3], a modified version of DBSCAN (iterative and multi-attribute) was used to cluster the different zones in the port, with the objective of improving organization and solving port congestion. In [37], they propose constructing the spatio-temporal similarity matrix using the dynamic time warping (DTW) algorithm from the data extracted by LPRs and applying DBSCAN to group different displacement patterns on the similarity matrix.…”

Exploring Vehicle Behavior in Smart Villages Environments: A Comprehensive Clustering Pipeline for Analysis and Insights

“…Few works related to clustering analysis in mobility use LPR cameras as the main source of information [73]. For example, [73] analyze the commuting patterns constructing the spatio-temporal similarity matrix using the dynamic time warping (DTW) algorithm; and afterward, analyze the characteristics of commuting patterns with the density-based spatial clustering of applications with noise (DBSCAN) algorithm. Similarly, [74] analyzes the change in traffic patterns during the pandemic using K-Means.…”