Abstract: The freeway traffic flow model is the foundation of designing, analyzing, simulating and operating the Traffic Surveillance&Control System of freeway. Due to the freeway traffic flow model is a high order nonlinear&time variant system, it makes the state estimation of the model become very difficult. This article briefly describes the traffic flow macro dynamic model in a traditional analytic model and the state estimation principle of the Extended Kalman Filter (EKF). Then it discusses the application of the model in a road section which has three sections of equal length, one entry and one exit. The EKF is adopted to simulate the density and velocity of the freeway vehicle flow. The simulation result proves the practicability of this algorithm.