An improved harmony search algorithm for blocking job shop to minimize makespan

In this paper, Harmony Search is applied to the blocking job shop problem with makespan minimization. According to the characteristics of the considered problem, a decoding method is introduced to generate feasible solutions. A rule is proposed to improvise new harmonies. Some approaches are developed to determine the harmony search considering rate, the pitch adjusting rate, the dynamic harmony memory. A local search is investigated to further improve quality of the solutions. Results of numerical experiments on classical benchmark instances show that the proposed algorithm can improve makespan 16.50% on average.