摘要： ► The proposed graph-pair representation technique for the unequal-area facility layout problem. ► Allows easy and appropriate settings of the departments’ relative locations. ► Allows easy and appropriate modifications of the departments’ relative locations. ► Does not impose any structural constraints (e.g. bands, cuts) on the solution space. ► May be used with a variety of metaheuristics. Owing to its theoretical as well as practical significance, the facility layout problem with unequal-area departments has been studied for several decades, with a wide range of heuristic and a few exact solution procedures developed by numerous researchers. In one of the exact procedures, the facility layout problem is formulated as a mixed-integer programming (MIP) model in which binary (0/1) variables are used to prevent departments from overlapping with one another. Obtaining an optimal solution to the MIP model is difficult, and currently only problems with a limited number of departments can be solved to optimality. Motivated by this situation, we developed a heuristic procedure which uses a “graph pair” to determine and manipulate the relative location of the departments in the layout. The graph-pair representation technique essentially eliminates the binary variables in the MIP model, which allows the heuristic to solve a large number of linear programming models to construct and improve the layout in a comparatively short period of time. The search procedure to improve the layout is driven by a simulated annealing algorithm. The effectiveness of the proposed graph-pair heuristic is demonstrated by comparing the results with those reported in recent papers. Possible extensions to the graph-pair representation technique are discussed at the end of the paper. 出版者： Amsterdam: Elsevier B.V 出版日期： 2012-04-16 出處： European journal of operational research, 2012-04, Vol.218 (2), p.382-391 版權： 2011 Elsevier B.V. 版權： 2015 INIST-CNRS 版權： Copyright Elsevier Sequoia S.A. Apr 16, 2012 識別號： ISSN: 0377-2217 識別號： EISSN: 1872-6860 識別號： DOI: 10.1016/j.ejor.2011.10.052 識別號： CODEN: EJORDT
