[CMZ02] Variable neighborhood search for the optimization of cable layout problem
Revue Internationale avec comité de lecture :
Journal Journal of Intelligent Manufacturing (Kluwer),
vol. 13(5), 2002
motcle:
Résumé:
This paper deals with the optimization of cable layout, which is a problem encountered by Electricité de France (EDF) in power plant
design. The problem consists in finding optimized cable routes connecting a set of pairs of facilities, respecting shelvesÂ’ capacities and a
number of other constraints. We show that the problem is equivalent to optimizing the cost of multiple paths subject to capacity
constraints. This is a special case of the integer min-cost multicommodity flow problem. In this paper, we first give a description of the
cable layout problem (CLP) and a modelization of the most important constraints. We then propose to solve this problem using the
extension variable neighborhood decomposition search (VNDS) of the recent variable neighborhood search (VNS) metaheuristic. This
metaheuristic combines local search with systematic changes of neighborhood in the descent phases to escape from local optima. In the
cable layout problem, we define the kth neighborhood of a solution S as the set of solution S¢ which differs from S in k cable paths. The
obtained results are validated by bounds provided by Lagrangian relaxation and compared to results obtained by a tabu search.