1991 年 57 巻 542 号 p. 3364-3371
This paper aims to design a flexible GT cellular layout and discusses its methodology. The method integrates production planning and GT cellular layout during a long-run planning horizon. The integrated planning model is formulated as a mixed-integer program (MIP), which contains two types of integer programming problems : determining (1) the production quantity and (2) the timing of the adjustment for the GT cellular layout in a finite planning horizon with a dynamic demand situation. This decision problem is solved so as to minimize the sum of inventory holding costs, group setup costs, material handling costs and layout adjusting costs subject to the capacity constraint and the demand requirement. The Benders decomposition is used to solve the MIP. The obtained dual problem is solved by linear programming using large-scale programming techniques. The decision rule determining the decision variables in the dynamic demand situation is proposed by examining the properties of the optimal solution obtained. And a method is also presented to cope with the uncertainty by adjusting individual production rates from their initially planned levels on the rolling horizon of T periods. Finally, a numerical example is demonstrated to explain the analytical results.