Optimal neural feedback control applied to a problem of economic growth in freight transport market

The evolution of a nation-wide freight transport market is a complex phenomenon that can be described by means of a "ad-hoc" developed dynamic model. This model depends on a set of control variables (i.e., the percentage of carbon tax on the fuel cost, the operational cost coverages, and growth rates of the various transportation modes) that can be chosen in a suitable way so as to minimize a given cost function (e.g., carbon emissions, public and private costs, fuel consumption, etc.). The problem has been addressed by searching for a feedback control law that can be approximated by means of the combination of both dynamic programming and neural networks. Preliminary simulation results with the afore-mentioned model are presented to demonstrate the effectiveness of the proposed method.