An enhanced memetic algorithm for energy-efficient and low-carbon flexible job shop scheduling problem considering machine restart

Abstract

The technological advancements of Industry 5.0 place greater emphasis on environmental sustainability and resilience for production scheduling. The flexible job shop scheduling problem (FJSP) effectively adapts to complex production environments and diverse scheduling requirements, which has made it an essential tool for studying modern production scenarios. Against this backdrop, this paper proposes an energy-efficient and low-carbon flexible job shop scheduling problem considering machine restart (ELFJSP-MR), aiming to minimize the makespan and total carbon emissions of the system. To solve ELFJSP-MR, we present an enhanced memetic algorithm (EMA) and design machine restart strategy to balance energy consumption and equipment lifespan. A multi-population hybrid model initialization based on logistic population growth model is used to enhance initial population diversity. Two novel neighborhood search methods are developed to improve convergence speed and explore the solution space more thoroughly. To enhance the flexibility and efficiency of local search, an adaptive operator selection model is designed. Finally, EMA and four well-known algorithms are evaluated on various benchmark problem instances. Experimental results demonstrate that EMA achieves faster convergence and greater stability for ELFJSP-MR. Furthermore, EMA exhibits exceptional performance across eight instances of aerospace composite material processing.