A flexible,adaptive traffic network simulation with parameter estimation

This article deals with macroscopic traffic modeling and online parameter estimation suitable for real-time simulation of complex road configurations. A numerical model for the nonlinear hyperbolic transport equation is introduced that works with a fundamental diagram of arbitrary shape and piecewise differentiable, discontinuous initial conditions. Suitable boundary conditions at road inlets and outlets are realized. A method to identify parameters of the underlying fundamental diagram via aggregated traffic sensor data is presented and the Fisher Information Matrix is utilized to optimize traffic sensor placement. The choice of parameterization, sensor placement, and the range of sensor data affect identifiability of the parameters of the fundamental diagram. The results are validated by comparison with a microscopic traffic simulation based on a car-following model.