Effective hybrid discrete artificial bee colony algorithms for the total flowtime minimization in the blocking flowshop problem

Abstract

In this paper, three effective hybrid discrete artificial bee colony (hDABC1, hDABC2, hDABC3) algorithms are presented to solve the blocking flowshop scheduling problem with the objective of minimizing the total flowtime. The three hybrid DABC algorithms utilize discrete job permutations to represent food sources and apply discrete operators to generate new food sources for the employed bees, onlookers, and scouts, respectively. First, two heuristic rules called the MME-A and MME-B (variant of combination of minmax and NEH) are presented to construct an initial population with a certain level of quality and diversity. Second, a self-adaptive strategy is applied to employed bees. Third, the estimation of distribution algorithm implements explicit learning from selected individuals and then generates good solutions for onlooker bees. Last but not least, to improve the algorithms' local exploitation ability, a very efficient local search-based insertion neighborhood is carried out in three stages respectively, that is, hDABC1 algorithm is generated by applying a local search to the solution obtained in the employed bee stage. hDABC2 is designed by carrying out a local search in the onlooker bee stage, and hDABC3 is developed by applying a local search in the scout bee stage. Computational experiments on standard benchmark problems are conducted. The results and comparisons show that the proposed algorithms are very effective and efficient for the blocking flowshop scheduling problems with total flowtime criterion than the other algorithms.