To MSA or not MSA? System convergence in the Danish national model

Abstract

In the paper we investigate the system convergence between the demand and supply system in a large scale transport model for Denmark. The demand is represented by the people that travel in the transport system and on the defined network, whereas the supply is represented by the level-of-service attributes (e.g., travel time and cost) offered to travellers.

For large scale models an “internal approach”, where the traffic assignment is integrated with the demand model, is not in practical possible. As a result we need to carefully consider the convergence of the “external loop” between the two models towards a common equilibrium. As iterations in the external loop are very costly from a computational point of view, there is a good payoff if convergence speed can be improved. Or put in even stronger words, if a sensible convergence cannot be obtained within relative few iterations, the model will be close to useless.

The paper is a continuation of work presented in Rich et al. (2010) based on a toynetwork. In this paper we investigate large scale performance in the newly implemented Danish National Model with 907 zones, more than 15 travel segments and state-of-the-art stochastic user equilibrium assignment models. In the paper we compare the method of repeated approximations (MRA) with different variants of the Method of Successive Averages (MSA).