EFJS']JSP-IBDR: Energy-efficient flexible job shop scheduling under incentive-based demand response via graph reinforcement learning with dual attention mechanism

被引:1
作者
Liu, Mingzhou [1 ]
Rui, Zhangjie [1 ]
Zhang, Xi [1 ]
Ge, Maogen [1 ]
Ling, Lin [1 ,2 ]
Wang, Xiaoqiao [1 ]
Liu, Conghu [3 ]
机构
[1] Hefei Univ Technol, Sch Mech Engn, Hefei 230009, Peoples R China
[2] Hefei Univ Technol, Anhui Key Lab Digital Design & Mfg, Hefei 230009, Peoples R China
[3] Suzhou Univ, Sch Mech & Elect Engn, Suzhou 234000, Peoples R China
基金
国家自然科学基金国际合作与交流项目; 中国国家自然科学基金;
关键词
Flexible job shop scheduling; Energy-efficient manufacturing; Incentive-based demand response; Graph reinforcement learning; Dual attention mechanism; OPTIMIZATION; MOEA/D;
D O I
10.1016/j.eswa.2025.127340
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
As energy demand grows and the penetration of intermittent renewable energy, optimizing energy consumption patterns in industrial production processes has become an increasing challenge for sustainable manufacturing. This study investigates an energy-efficient flexible job shop scheduling problem with incentive-based demand response (EFJSP-IBDR), aiming to enhance energy consumption efficiency and cost-effectiveness. Given the time-sensitive and complex state features of EFJSP-IBDR, we propose an end-to-end graph reinforcement learning (GRL) approach. Initially, we decompose EFJSP-IBDR into a set of preference-based subproblems to facilitate multi-objective optimization. On this basis, a Markov decision process (MDP) with heterogeneous graph states is employed to model these subproblems. Subsequently, we construct graph neural network schedulers that integrate dual attention mechanisms to extract the features necessary for decision-making in each subproblem. Additionally, a similarity-based parameter transfer strategy is designed to accelerate the training process of all schedulers. Empirical results demonstrate that our approach can reduce peak demand by 47.22% and average load by 51.01% during the manufacturing cycle. The effectiveness of the proposed method suggests its potential to improve load-side energy consumption flexibility and grid resilience.
引用
收藏
页数:18
相关论文
共 66 条
[1]   Scheduling for sustainable manufacturing: A review [J].
Akbar, Muhammad ;
Irohara, Takashi .
JOURNAL OF CLEANER PRODUCTION, 2018, 205 :866-883
[2]   A multi-site production and microgrid planning model for net-zero energy operations [J].
An Pham ;
Jin, Tongdan ;
Novoa, Clara ;
Qin, Jin .
INTERNATIONAL JOURNAL OF PRODUCTION ECONOMICS, 2019, 218 :260-274
[3]   An improved non-dominated sorting biogeography-based optimization algorithm for the (hybrid) multi-objective flexible job-shop scheduling problem [J].
An, Youjun ;
Chen, Xiaohui ;
Li, Yinghe ;
Han, Yaoyao ;
Zhang, Ji ;
Shi, Haohao .
APPLIED SOFT COMPUTING, 2021, 99
[4]  
[Anonymous], 2024, CO2 Emissions in 2023
[5]  
Brandimarte P., 1993, Annals of Operations Research, V41, P157, DOI 10.1007/BF02023073
[6]   A Pareto based discrete Jaya algorithm for multi-objective flexible job shop scheduling problem [J].
Caldeira, Rylan H. ;
Gnanavelbabu, A. .
EXPERT SYSTEMS WITH APPLICATIONS, 2021, 170
[7]   Demand response in US electricity markets: Empirical evidence [J].
Cappers, Peter ;
Goldman, Charles ;
Kathan, David .
ENERGY, 2010, 35 (04) :1526-1535
[8]   Demand-Side Regulation Provision From Industrial Loads Integrated With Solar PV Panels and Energy Storage System for Ancillary Services [J].
Chau, Tat Kei ;
Yu, Samson Shenglong ;
Fernando, Tyrone ;
Iu, Herbert Ho-Ching .
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2018, 14 (11) :5038-5049
[9]   A Deep Reinforcement Learning Framework Based on an Attention Mechanism and Disjunctive Graph Embedding for the Job-Shop Scheduling Problem [J].
Chen, Ruiqi ;
Li, Wenxin ;
Yang, Hongbing .
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2023, 19 (02) :1322-1331
[10]   Energy-aware operations management for flow shops under TOU electricity tariff [J].
Cui, Weiwei ;
Lu, Biao .
COMPUTERS & INDUSTRIAL ENGINEERING, 2021, 151