This project aims to monitor traffic stream and identify vehicle types, vehicle trajectories, volume, occupancy and speed, and based on which to model microscopic lead-vehicle behavior, vehicle-following behavior and lane-changing behavior, and to build intelligent agent-based traffic flow simulation at macroscopic level with high fidelity. The project improves traffic operations efficiency and safety, reduces carbon emission, and enhances air quality by optimizing operations policy, regularization and signal.