Optimal routing in freeway networks via sequential linear programming

Reference:
Z. Cong, B. De Schutter, and R. Babuska, "Optimal routing in freeway networks via sequential linear programming," Proceedings of the 10th IEEE International Conference on Networking, Sensing and Control, Paris, France, 6 pp., Apr. 2013. Paper FrB01.5.

Abstract:
Based on the Ant Colony Optimization (ACO) algorithm, we previously developed an optimization method to solve the dynamic traffic routing problem in freeway networks, called Ant Colony Routing (ACR). This method uses virtual ants to search appropriate routes in a virtual ant network, and accordingly distributes the vehicles over the corresponding traffic network sharing the same topology with the ant network. By using Model Predictive Control (MPC), we can iteratively apply ACR at each control step to generate a control signal - i.e. splitting rates at each node in the traffic network. Motivated by the MPC framework with ACR, we show in this paper that sequential linear programming (SLP) can be used as optimization method for solving the dynamic traffic routing problem in some specific cases, resulting a lower computation time while achieving a similar performance as the ACR algorithm.