Spatial-temporal traffic congestion identification and correlation extraction using floating car data

Abstract

Traffic congestion induces significant economic loss each year, and identifying traffic congestion patterns is necessary for better traffic control and management. Floating car data (FCD) provides a cost-effective alternative for assessing traffic status and detecting congestion on a large scale. Recently, a new speed performance index (SPI) has been proposed to evaluate traffic status considering both traffic flow speeds and road speed limits. This research proposes a new categorization criterion to define traffic conditions as five levels based on SPI values, and applies the proposed criterion in a case study to investigate traffic status and detect traffic congestion patterns on urban freeways based on FCD analysis. The research results reveal a clear understanding regarding urban freeway traffic status and congestion spatial and temporal distribution. With more comprehensive FCD from larger spatial and temporal domains, continued research could focus on travel speed prediction, travel time estimation, prediction, and reliability analysis at both the road segment and network levels.