A new hybrid algorithm to solve the cell formation problem considering multiple process routings

In this study, a hybrid algorithm with Particle Swarm Optimization and Competitive Neural Network is proposed to solve the manufacturing cell formation problem considering multiple process routings for parts. Alternate process routings are selected by Particle Swarm Optimization, after that Competitive Neural Network is working on composing the cells using these selected routes. The main objective of this algorithm is maximizing the Grouping Efficacy. Specific side of this study is particles that inside the PSO method searching in unit hypercube solution space with continuous values. Continuous values are transformed to discrete values for obtaining alternate routes. This approach is tested on the most known problems in literature and compared with other methods. By these comparisons, it is comprehended that this new approach reaches better solution for some of these problems.