Building on the research performed in Phases I and II of this study, the authors develop in Phase III an algorithm for estimating vehicle turning proportions at signalized intersections using partial detector information. This report demonstrates the efficiency of a maximum likelihood estimator approach to estimate turning proportions over a prediction error minimization principle. This phase enhances the traffic flow models to provide more detail on short-term instability and left-turning behavior at intersections. Finally, the authors develop two types of estimation algorithms using Nonlinear Least Squares and Quasi Maximum Likelihood to estimate the turning proportion parameters. These algorithms are useful tools to generate real-time applications in traffic control.
Davis, Gary A.; Stephanedes, Yorgos J.; Lan, Chang-Jen.
Real-Time Traffic Prediction for Advanced Traffic Management
Systems: Phase III.
Retrieved from the University of Minnesota Digital Conservancy,
Content distributed via the University of Minnesota's Digital Conservancy may be subject to additional license and use restrictions applied by the depositor.