Traffic Flow Estimation Based on Toll Ticket Data Considering Multitype Vehicle Impact

Abstract

Traffic flow estimation (TFE) plays an important role in transportation systems, especially in relation to the detection, planning, and management of traffic for expressway systems. However, traditional approaches to TFE using roadside detectors usually have low-coverage and high-cost problems. This paper proposes a toll ticket data (TTD)–based approach to TFE for filling the gaps in existing approaches and also considers multitype vehicles’ impact on traffic flow for more accurate TFE results. First, route choice models are constructed to capture travel characteristics of mult-type vehicles and acquire route traffic flow from TTD. Then, a simulation-based dynamic traffic assignment (DTA) algorithm is developed to obtain traffic flow regarding vehicle type, time, and location, with the consideration of multitype vehicles’ impact by using a tollbooth conversion factor (TCF) and point-of-queue model (PQM). The proposed approach is evaluated empirically using data relating to the Shandong Expressway Toll System in China. The results confirm that it performs effectively. This research could contribute to future TFE and provide references for dynamic, delicate traffic management in an economical and reliable way.