Abstract
In this paper, we propose new memetic algorithms (MAs) for the multiobjective flexible job shop scheduling problem (MO-FJSP) with the objectives to minimize the makespan, total workload, and critical workload. The problem is addressed in a Pareto manner, which aims to search for a set of Pareto optimal solutions. First, by using well-designed chromosome encoding/decoding scheme and genetic operators, the nondominated sorting genetic algorithm II (NSGA-II) is adapted for the MO-FJSP. Then, our MAs are developed by incorporating a novel local search algorithm into the adapted NSGA-II, where some good individuals are chosen from the offspring population for local search using a selection mechanism. Furthermore, in the proposed local search, a hierarchical strategy is adopted to handle the three objectives, which mainly considers the minimization of makespan, while the concern of the other two objectives is reflected in the order of trying all the possible actions that could generate the acceptable neighbor. In the experimental studies, the influence of two alternative acceptance rules on the performance of the proposed MAs is first examined. Afterwards, the effectiveness of key components in our MAs is verified, including genetic search, local search, and the hierarchical strategy in local search. Finally, extensive comparisons are carried out with the state-of-the-art methods specially presented for the MO-FJSP on well-known benchmark instances. The results show that the proposed MAs perform much better than all the other algorithms.