Une Approche Dynamique pour l'Optimisation des Trajectoires de n Véhicules à Guidage Automatique

摘要

We propose an analytic method based on dynamic programming for determining optimal trajectories in a rather delicate problem of transport: by Automated Guided Vehicles (AGV). These are driverless electric vehicles used in many fields including industry. Langevin et al. (Int J Flex Manuf Syst 8(3):247-262, 1996) showed how to determine the optimal solution in the case of two vehicles. Our method generalizes them with some modifications to not generate an exorbitant number of states. For a given state and a mission we offer a single successor instead of n and we prove that the states, in considerable number, that we exclude are not in the optimal trajectory.