Hybrid Approach for Solving Multi-Objective Hybrid Flow Shop Scheduling Problems with Family Setup Times

Abstract

Solving industrial scheduling problems remains challenging despite the heavy research efforts in the last decade due to the introduction of new technologies in the context of industry 4.0. Such problems must be solved with light execution time to support near-real-time decision-making processes. In addition, the majority of real industrial Hybrid Flow Shop (HFS) scheduling problems must be solved to minimize multi-objective values that are conflicting in nature. Such problems are proven to be NP-hard. Therefore, in this paper, a hybrid multi-objective approach is presented for solving HFS scheduling problems. The hybrid technique is based on the Non-Dominated Sorting Genetic Algorithms three (NSGA III). The design of the hybrid approach is based on the use of multi-populations that are used to control different allocation-, sequencing-algorithms, and decision points over time. The problems are solved to minimize the makespan, the total number of family major setup times, the total tardiness, and the total number of penalties. The computational results show that the presented approach marginally outperforms other techniques that are published in the related literature for solving two-stage and four-stage HFS scheduling problems. Both problems are derived from real use-cases with real problem instances.