Congestion Scenario-based Vehicle Classification Detection Models Based on Traffic Flow Characteristics and Observed Event Data

While the existing applied length-based vehicle classification model has been to estimate vehicle lengths accurately with dual-loop traffic monitoring station data under free traffic condition, it produces considerable errors against congested traffic. In this study, both ground-truth vehicle trajectory and simultaneous loop event data are used to characterize the impact of congested traffic on vehicle classification. Eight scenarios are synthesized to define the vehicles’ stopping locations over two single loops of the dual-loop station. Under the synchronized traffic flow, acceleration or deceleration is considered in the new developed Vehicle Classification under Synchronized Traffic Model (VC-Sync model) to reflect the speed variation between loops. As a result, the error of the vehicle classification is reduced from 33.5% to 6.7%, compared to the existing applied model. Under the stop-and-go traffic condition, a Stop-on-Both-Loops-only (SBL) was developed along with the VC-Sync model to simplify the complexity of congested traffic situation in vehicle length estimation. The error is reduced by using the SBL model from 235% to 17.1%, compared to the existing applied model. Capability of identifying traffic phases is a critical prerequisite to applying the new vehicle classification models under congestions. An innovative method for identifying the traffic phases has been therefore proposed based on the existing traffic stream models along with the new findings of the authors’ empirical data analysis. As a result, a heuristic traffic phase identification model has developed and successfully applied in the case study for evaluating the new length-based vehicle classification models with dual-loop data. Wei, Ai, Liu, Li, and Wang 1