A Data-Driven Model for the Energy-Efficient No-Wait Flexible Flow Shop Scheduling Problem with Learning and Deteriorating Effects

This work aims to investigate an energy-efficient no-wait flexible flow shop problem considering deteriorating and learning effects under uncertainty. To do this, a data-driven decision-making framework is developed in this research. At the outset, a multi-objective mathematical model is proposed for the research problem that minimizes the makespan, total tardiness, and total energy consumption. Then, to tackle uncertainty, a data-driven approach based on the fuzzy robust optimization, Seasonal Autoregressive Integrated Moving Average and Support Vector Regression methods is developed. Afterwards, to solve the proposed model, a hybrid approach based on the LP-Metric method and metaheuristic algorithms is proposed. The achieved outputs confirm the appropriate performance of the developed data-driven approach. Based on the obtained results, the developed hybrid metaheuristic algorithm shows an appropriate performance in both computational time and solution quality metrics. Also, the outputs indicate that the objective functions of the proposed model have increased when the due date parameter increases. Additionally, results show that with the increase in the absolute value of the learning coefficient, the first, second, and third objective functions of the model have decreased.