Towards system-aware routes

Abstract

We introduce solution methods for finding a system-optimal or at least system-aware set of routes for multiple concurrently travelers on a road network such that congestion is minimized, i.e. traffic flow is maximized. Beside a dynamic programming approach, we present a basic heuristic mimicking the current state-of-the-art resulting in a user equilibrium and a multistart local search based approach approximating the system-optimum. Experimental results show that the chosen approaches are promising and that heavy congestion can be avoided when applying such a planning approach.