Neural approximations for feedback optimal control of freeway systems

The problem of clearing congestion situations in freeway traffic is address ed for both an N-stage and an infinite-stage control horizon (in the latter case, a receding-horizon control mechanism is used). Traffic is controlled by regulating the vehicle access to the freeway and by limiting the vehicl e speed by means of variable message signs. To describe the traffic behavio r, a "classical" macroscopic model, first proposed by Payne, is adopted. Ev en though the problem is stated within a deterministic context, an optimal control law in feedback form is sought to react to unpredictable events. Th e resulting functional optimization problem is reduced to a nonlinear progr amming problem by constraining the control law to take on a tired structure in which free parameters have to be optimized. For such a structure, a mul tilayer feedforward neural mapping is chosen. Simulation results show the e ffectiveness of the proposed method in two different case studies. For the simulation of the second case study, real traffic data are used, which allo ws one to very well represent critical traffic conditions on freeways.