A grouping genetic algorithm for the multi-objective cell formation problem

Abstract

In this research, we propose an efficient method to solve the multi-objective cell formation problem (CFP) partially adopting Falkenauer's grouping genetic algorithm (GGA). The objectives are the minimization of both the cell load variation and intercell flows considering the machines’ capacities, part volumes and part processing times on the machines. We relax the cell size constraints and solve the CFP without predetermination of the number of cells, which is usually difficult to predict in a real-world CFP design. We also make some effort to improve the efficiency of our algorithm with respect to initialization of the population, fitness valuation, and keeping crossover operator from cloning. Numerical examples are tested and comparisons are made with general genetic algorithms (GAs). The result shows that our method is effective and flexible in both grouping machines into cells and deciding on the number of cells for the optimal solution.