Verified feasibility of structured multicommodity flow solutions
(English)Manuscript (preprint) (Other academic)
This paper introduces a new approach for verifying feasibility for multicommodity flow problems, MFP:s for short. This feasibility problem priginates in a new solution method for the convex MFP based on the solution of the dual convex MFP, where optimality is demonstrated by showing that a given solution to the dual convex MFP is feasible for the convex MFP and hence optimal. In this paper a brief description of the MFP and the structure of its solutions are given. Furthermore, a distance minimizing method, based on simplicial decomposition, is described and used to efficiently verify feasibility and hence demonstrate optimality. The method is applied to convex MFP:s arizing in a traffic assignment setting.
Engineering and Technology
IdentifiersURN: urn:nbn:se:liu:diva-85562OAI: oai:DiVA.org:liu-85562DiVA: diva2:571688