Energy-efficient bi-objective manufacturing scheduling with intermediate buffers using a three-stage genetic algorithm

Abstract

The flexible multi-task scheduling problem has been extensively investigated in manufacturing systems, and its objectives are often related to the quality of manufacturing services. However, energy-related objectives along with workload balance have rarely been considered. Thus, a novel bi-objective optimization model is proposed to achieve green manufacturing. The Pareto-based fitness evaluation is employed to make a trade-off between total energy consumption and workload balance. Intermediate buffers are also considered, making the model more practical and more complicated. To solve the proposed model, a new three-stage genetic algorithm (3S-GA) is presented. A Pareto-based adaptive population size method is proposed to maintain the diversity of the population and ensure the convergence rate. To cope with the subtask sequencing complexity, a real-time sequence scheduling heuristic is explored to effectively initialize the subtask sequence to save the energy in manufacturing systems, which is designed by minimizing the standby time according to the laxity of subtasks. After a series of experimental designs based on the Taguchi method, a suitable parameter combination of the 3S-GA is utilized. Further, computational experiments based on five instances demonstrate that the 3S-GA outperforms other four baseline algorithms taken from the literature in solving the proposed bi-objective optimization model.