A dynamic programming–enhanced simulated annealing algorithm for solving bi-objective cell formation problem with duplicate machines

• Main Authors: Mohammad Mohammadi, Kamran Forghani
• Format: Article
• Language: English
• Published: Growing Science 2015-04-01
• Series: Decision Science Letters
• Subjects: Cellular manufacturing; Hybrid simulated annealing; Dynamic programming; Machine duplication; Job shop; Flow shop
• Online Access: http://www.growingscience.com/dsl/Vol4/dsl_2014_39.pdf

Cell formation process is one of the first and the most important steps in designing cellular manufacturing systems. It consists of identifying part families according to the similarities in the design, shape, and presses of parts and dedicating machines to each part family based on the operations required by the parts. In this study, a hybrid method based on a combination of simulated annealing algorithm and dynamic programming was developed to solve a bi-objective cell formation problem with duplicate machines. In the proposed hybrid method, each solution was represented as a permutation of parts, which is created by simulated annealing algorithm, and dynamic programming was used to partition this permutation into part families and determine the number of machines in each cell such that the total dissimilarity between the parts and the total machine investment cost are minimized. The performance of the algorithm was evaluated by performing numerical experiments in different sizes. Our computational experiments indicated that the results were very encouraging in terms of computational time and solution quality.