تخمين زن تطبيقي حالات ترافيك با استفاده از ايده تركيب تخمين پارامترهاي مدل ترافيك

Adaptive Freeway Traffic State Estimator based on Fusion of the Model Parameters Estimates

Abstract: This paper presents real-data testing results of a real-time freeway traffic state estimator. The general approach to real-time adaptive of freeway traffic state estimation is based on nonlinear macroscopic traffic flow modeling and extended Kalman fliter algorithm. Macroscopic traffic flow model contains three important and unknown parameters (free speed, critical density and exponenet), which should be estimated with off-line or on-line methods. One innovative aspects of the estimator is the real-time joint estimation of traffic flow variables (traffic flow, mean speed and traffic density) and model parameters, that leads to some significant features such as: avoidance of prior model calibration, automatic adaption to changing external conditions (e.g. weather conditions, traffic composition,...). The purpose of the reported real-data testing is, first, to demonstrate some obstacles in previous methods, second, to propose two methods based on dual filtering and fosion of the model parameter estimates for improving the previous methods.