Multi-objective integrated scheduling of ship pipe hybrid flow shop with heterogeneous crane transportation

In ship pipe production workshops, frequent crane-based material handling is required, resulting in strong coupling between processing and transportation. By considering these practical characteristics, a hybrid flow shop scheduling problem with coordinated transportation of heterogeneous cranes is defined. Accordingly, a mixed-integer programming model is developed to minimize makespan, total energy consumption (TEC), and total carbon emissions (TCE). An adaptive learning multi-objective evolutionary algorithm (ALMOEA) is then proposed, integrating a multi-layer matrix encoding scheme, a first-come-first-served-based right-shift partition decoding mechanism, differentiated evolutionary strategies, and a Q-learning- and fuzzy correlation entropy-driven adaptive local search module. Comprehensive experiments are conducted on 30 benchmark instances by comparing ALMOEA with four representative algorithms. The results indicate that ALMOEA achieves better Hypervolume (HV) and Inverted Generational Distance (IGD) performance, which is further supported by Kruskal-Wallis tests followed by Dunn's post-hoc tests with Holm correction. Moreover, runtime analysis reveals that ALMOEA maintains a favorable trade-off between solution quality and computational overhead. Finally, a real-world case validates its practical applicability, with reductions of approximately 10% – 30% in makespan and approximately 3% – 7% in both TEC and TCE compared with competing algorithms. Overall, ALMOEA effectively improves scheduling efficiency while reducing environmental impact in ship pipe production.