Search

This study aims to investigate the congestion of the urban business circle based on traffic performance index (TPI) data. A hierarchical clustering algorithm is adopted for the data analysis. A dataset of nearly 64,260 pieces of TPI data from July to October 2017 in Qingdao is collected, and its features, such as interval characteristics, mean time distribution and spatiotemporal correlation, are analyzed. The results show that the southern coastal commercial circle of Qingdao is more congested than the other circles; there exists a morning and afternoon peak, with two peaks on workdays; otherwise, weekend and vacation periods do not show congestion. The congestion level toward the end of the vacation week (October 4th–8th) is lower than that during the beginning (October 1st–3rd). Considering the temporal and spatial dimensions, the causes of the two different congestion states during the holidays are speculated upon The characteristics of traffic congestion in Qingdao business circle and its possible causes are proposed, and a new design and overall plan for Qingdao is promoted. • Traffic congestion characteristics of business circle are extracted from TPI data. • The agglomerative hierarchical clustering algorithm is adopted to analyse data. • Trends and patterns of traffic congestion are summarized temporarily and spatially. [ABSTRACT FROM AUTHOR]