Accuracy Analysis of Freeway Traffic Speed Estimation Based on the Integration of Cellular Probe System and Loop Detectors

Traffic speed estimation is a fundamental task for traffic management centers and a critical element of intelligent transportation systems. For this purpose, various sensors are used to collect traffic speed information. The cellular probe system is gaining market penetration and becomes a newly effective and practical traffic speed measurement technique. In this article, the handoff system, one type of cellular probe technology, is used for speed detection. However, the handoff coverage size is usually variable and consecutive handoff points are usually far apart on a freeway. To improve the accuracy of speed estimation, this article proposes a travel-time-based method to aggregate the estimation results of the cellular probe system and loop detectors. For the purpose of a rigorous analysis, data are generated from microscopic simulation models of virtual one-direction freeway segments under various traffic conditions. Thus, the correlations between estimation accuracy and handoff distance, traffic condition, and the number of loop detectors are evaluated and analyzed. The results show that best performance is achieved with the shortest handoff distance. The aggregation of estimated speeds from cellular probe system and loop detectors can improve the speed estimation accuracy by taking advantage of each sensor if the space of loop detectors is more than 500m. Also, an increasing number of loop detectors will improve the accuracy. Furthermore, the improvement of integration accuracy is much better under free flow conditions than under congested conditions.