A generic optimization framework for scheduling problems under machine deterioration and maintenance activities. (December 2022)

• Record Type: Journal Article
• Authors: Rudek, Radosław
• Is Part Of: Computers & industrial engineering. Volume 174(2022)
• Publication Date: 2022-12
• Subjects: Scheduling -- Maintenance -- Industry 4.0 -- Representation -- Robust -- Metaheuristic; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2022.108800
• Languages: English
• ISSNs: 0360-8352

Abstract

Abstract: This paper introduces a novel generic optimization framework for scheduling problems with machine deterioration and maintenance activities. The framework employs computationally efficient and robust data structures for schedules along with templates for implementing optimization algorithms making it applicable within Industry 4.0. The concept includes, but not limited to, the optimization objectives such as the maximum completion time, the maximum lateness (or tardiness), the total (weighted) number of late jobs, the total (weighted) completion times, the total (weighted) tardiness, the just-in-time. A set of example algorithms based on iterative local search, i.e. Nawaz–Enscore–Ham's method (NEH), simulated annealing, genetic algorithm, scatter search algorithm, artificial bee colony were embedded within the framework to demonstrate its robustness. The theoretical and experimental analysis presented in the paper proved the high efficiency of the proposed approach; it traverses areas containing optimal solutions and at the same time it is characterized by a low computational complexity of related procedures to calculate criterion values or to generate new solutions. Furthermore, if the optimization methods are enhanced by the proposed data structures and procedures, their standard computational complexity is not increased even for complex scheduling problems including maintenance activities. The proposed framework modular architecture enables further integration with new algorithms aiming at improvement of neighbourhood search techniques. Its robustness, efficiency, and modularity can bring advantages in various decision support systems. Highlights: A computationally efficient generic optimization framework is constructed. Heuristic and metaheuristic optimization algorithms are enhanced by the framework. Scheduling problems under deteriorating and maintenances are modelled. The experiments confirm that our approach is computationally efficient and robust.