Work

Energy-efficient flexible job shop scheduling under time-of-use electricity pricing using a reinforcement learning-assisted evolutionary algorithm

Year: 2026
Type: article
Abstract: To address the difficulty of jointly optimizing production efficiency and electricity cost in flexible job shop scheduling under time-of-use electricity pricing, a bi-objective scheduling model is established to minimize the makespan and total electricity cost. An energy-aware hybrid reinforcement optimization algorithm (EAHROA) is proposed. The algorithm adopts a two-segment OS-MS encoding scheme to represent operation sequencing and machine selection. During decoding, a time-of-use electricity price-aware mechanism is introduced to calculate processing and idle electricity costs, guiding operations to avoid peak-price periods when possible. Three neighborhood structures, namely critical-machine reassignment, tariff-aware neighborhood search, and critical-operation insertion/swap, are designed to enhance local optimization for critical operations and high-cost operations. Furthermore, a Q-learning controller selects a joint search policy, including the crossover probability, mutation probability, neighborhood type, and maximum waiting limit, according to the search stage, the proportion of solutions in the first Pareto front, and the recent search trend. Simulation experiments on Brandimarte benchmark instances show that EAHROA provides competitive trade-offs between makespan and total electricity cost and achieves improved IGD, HV, and SC results in the comparative studies.
Language: English
Open Access status: gold
Estimated APC paid: $2,690
OpenAlex ID: W7226024100
FWCI:
Cites:
Cited by:
Related to: 10