Application of Kalman Filtering Method Based on Real-time OD Estimation

OD estimation is a critical component of Dynamic Traffic Assignment System in that it determines the frequencies of driver´s trips through a network. As OD estimation is intended for use in a real-time setting, computational issues are critical, and the Kalman Filtering method is provided to increase computational efficiency. The basic idea of this applied is to use all the information contained in historical OD data in conjuction with data on traffic counts to generate OD estimated in real-time. Unlike other approaches in literatures, this approach is based on deviations from historical values.