In this paper we study a minimum cost, multicommodity network flow problem in which the total cost is piecewise linear, concave of the total flow along the arcs. Specifically, the problem can be defined as follows. Given a directed network, a set of pairs of communicating nodes and a set of available capacity ranges and their corresponding variable and fixed cost components for each arc, the problem is to select for each arc a range and identify a path for each commodity between its source and destination nodes so as to minimize the total costs. We also extend the problem to the case of piecewise nonlinear, concave cost function. New mathematical programming formulations of the problems are presented. Efficient solution procedures based on Lagrangean relaxations of the problems are developed. Extensive computational results across a variety of networks are reported. These results indicate that the solution procedures are effective for a wide range of traffic loads and different cost structures. They also show that this work represents an improvement over previous work made by other authors. This improvement is the result of the introduction of the new formulations of the problems and their relaxations.
[1]
W. Zangwill.
Minimum Concave Cost Flows in Certain Networks
,
1968
.
[2]
Philip Wolfe,et al.
Validation of subgradient optimization
,
1974,
Math. Program..
[3]
Richard M. Soland,et al.
Optimal Facility Location with Concave Costs
,
1974,
Oper. Res..
[4]
Bezalel Gavish,et al.
On obtaining the 'best' multipliers for a lagrangean relaxation for integer programming
,
1978,
Comput. Oper. Res..
[5]
Warrren B Powell,et al.
The load planning problem of motor carriers: Problem description and a proposed solution approach
,
1983
.
[6]
Warren B. Powell,et al.
A Local Improvement Heuristic for the Design of Less-than-Truckload Motor Carrier Networks
,
1986,
Transp. Sci..
[7]
Clyde L. Monma,et al.
Send-and-Split Method for Minimum-Concave-Cost Network Flows
,
1987,
Math. Oper. Res..
[8]
Stephen C. Graves,et al.
A composite algorithm for a concave-cost network flow problem
,
1989,
Networks.