A stochastic traffic assignment algorithm based on Ant Colony Optimisation

In this paper we propose a Stochastic User Equilibrium (SUE) algorithm that can be adopted as simulation model in the design problem or in the real-time management of transportation systems when cost functions (i.e. travel times) depend on number of vehicles that utilises network elements (congestion assumption). Indeed, analyses of real dimension networks need of simulation algorithms that allow to determine network conditions and performances in short times. Hence, we developed an MSA (Method of Successive Averages) algorithm based on the Ant Colony Optimisation paradigm that allows to simulate transportation systems in shorter time but with the same accuracy respect to traditional MSA algorithms. Finally, by means of Blum's theorem, we stated theoretically the convergence of the proposed ACO-based algorithm.