Development and validation of an areawide congestion index using Intelligent Transportation Systems data

Abstract

The primary objective of this study was to examine potential areawide congestion indices using loop detector data, and to select one that best explains the congestion of a freeway system. Another objective was to determine the accuracy of the loop detector data. Seven potential congestion indices were calculated from TransGuide loop detector data for an entire month. These indices were evaluated to see which measures best fit the requirements of an effective areawide congestion index. From this research, it was determined that the travel rate ratio met all but one of the criteria of a successful index. Travel rate ratio is the ratio of congested travel time to travel time in free-flow conditions. The weakness of travel rate ratio is that it is an averaged value, so it cannot measure the extent (coverage) of the congestion. The average delay per vehicle in delay and congestion severity index met all but two of the criteria. Hourly traffic volume data from the TransGuide loop detector system and the Automatic Traffic Recorder (ATR) stations were checked for data accuracy. On average, the ATR traffic volumes were slightly more accurate than the TransGuide volumes. The volumes from two ATR stations were compared with nearby TransGuide loop detector volumes to evaluate the similarity of TransGuide and ATR volumes. It was concluded that even though the traffic volumes may be inaccurate, most of the loop detectors studied were strongly correlated with the ground truth due to consistent over-counting or under-counting.