Network-Wide Mesoscopic State Estimation Based on a Variational Formulation of the LWR Model and Using Lagrangian Observations

Abstract

This paper presents a generic data assimilation framework based on a mesoscopic-LWR model formulated in Lagrangian-space coordinates and using both Lagrangian and Eulerian observations. This is a challenging work since probe trajectories are not directly related to specific vehicle/platoon indexes in the simulation model. Therefore, we develop a method to incorporate probe information and to further estimate states. The proposed method has been validated on a homogeneous road stretch, and it provides promising results for further extension of the framework.