A multi-objective scheduling method for distributed and flexible job shop based on hybrid genetic algorithm and tabu search considering operation outsourcing and carbon emission
Wenxiang Xu,Yongwen Hu,Wei Luo,Lei Wang,Rui Wu
DOI: https://doi.org/10.1016/j.cie.2021.107318
2021-07-01
Abstract:<p>The general distributed flexible job shop scheduling problem (DFJSP) involves three sub-problems: (1) operation-sequencing, (2) job-to-cell assignment, and (3) operation-to-machine assignment. Due to the variety of operations involved in large-scale and complex workpieces, it is difficult for a single manufacturing enterprise to finish all operations of some workpieces. Therefore, in addition to the above three sub-problems, the characteristic that some workpieces need to carry out operation outsourcing was also considered in this study, so that the model established in this paper was more in accordance with the actual situation of manufacturing enterprises than the existing studies. Besides, considering the increasingly prominent environmental problems and the increasing costs of environmental protection, the mathematical model of this paper includes four optimization objectives: makespan, costs, quality and carbon emission. To highlight the importance of each objective and improve the solving efficiency of related algorithms, we used the fuzzy analytical hierarchy process (FAHP) to transform the multi-objective problem into a single objective problem. Aiming at the above model, a hybrid genetic algorithm and tabu search (H-GA-TS) with three-layer encoding was developed in this paper. The algorithm combined the advantages of both genetic algorithm (GA) in global search and tabu search (TS) in local search, and realized the performance improvement when compared with the above two simple algorithms. The comparison experiment conducted performed in this study also verified the effectiveness and performance advantages of the hybrid algorithm over GA and TS.</p>
computer science, interdisciplinary applications,engineering, industrial