Simulated annealing procedures for forming machine cells in group technology

The group technology (GT) problem deals with grouping parts with similar design and/or process characteristics into part families and forming associated machines into machine cells in order to achieve a higher level of manufacturing efficiency. Various models and solution approaches for solving the GT problem have been proposed in the literature. In this paper, we propose a quadratic programming model for forming machine cells so as to maximize the sum of machine similarities within cells while subject to cell size limitation. Our focus is on the development of simulated annealing based algorithms for solving the cell formation problem. A systematic computational test is conducted to test the performance of the proposed algorithms. Our computational results show the superiority of the simulated annealing algorithms over a graph partitioning based heuristic.