COAL: A new heuristic approach for solving the fixed charge problem-Computational results

Computational results are presented for a new heuristic algorithm for solving large general fixed charge problems based on Lagrangian relaxation and cost allocation heuristics; these results are contrasted with results using adjacent extreme point heuristics and branch and bound mixed integer programming. General facility location and capacitated warehouse location problems as well as production planning problems with fixed and variable work force considerations are used. The many test problems evaluated include problems with a large number of fixed charge and continuous variables. The new algorithm demonstrates superiority over alternative methods for solving these problems.