Network flow problems with one side constraint: A comparison of three solution methods (Q1102172)

We consider the minimum cost network flow problem subject to one side constraint. Three methods for solving such problems are presented and their computational efficiency is compared. The first method is a specialized primal simplex algorithm which exploits the underlying network structure. The second algorithm is a straight forward dual method which successively reduces the infeasibility of the side constraint. Finally, a Lagrangean approach is tested which uses a relaxation of the side constraint. Computational experience indicates that the specialized primal algorithm is superior to the other approaches for all but very small problems.