A bilevel descent method of optimizing signals and prices for a multi-modal network while taking account of travellers' choices (equilibrium) is specified within a framework which may, when developed, be efficient for large networks. Similar trilevel methods for the corresponding dynamic problem are also presented. Fairly complete proofs of convergence of the method to a local optimum are given for the steady state case but there remains a gap when the authors seek to prove convergence in a dynamic context; however, if the method converges (in a dynamic context) to the set of equilibria then (under natural conditions) it must also converge to the set of local optima.
|Title of host publication||Transportation Networks: Recent Methodological Advances|
|Subtitle of host publication||Selected Proceedings of the 4th EURO Transportation Meeting|
|Publication status||Published - 1998|