Mapping to cells: a map-independent approach for traffic congestion detection and evolution pattern recognition

Map matching is a fundamental prerequisite for traffic engineers in detecting congestion using location data represented by trajectory data. Previous studies often revolve around road matching, yet limitations arise from trajectory data quality and map-matching accuracy. This paper introduces a map-independent congestion identification method, involving urban cell network construction, congestion modeling with speed fluctuations, and the exploration of congestion evolution patterns. Finally, we validated our proposed method using Floating Taxi Data (FTD) from Xi'an, China. The result indicates that the method proposed in this study can identify urban traffic congestion and uncover its evolutionary characteristics without relying on maps. In contrast to other metrics, the customized congestion value considers the impact of speed fluctuations on congestion. The method proposed in this paper offers a benchmark solution for characterizing urban traffic congestion and formulating travel guidelines.