Leveraging constraint programming in a deep learning approach for dynamically solving the flexible job-shop scheduling problem

被引:0
|
作者
Echeverria, Imanol [1 ]
Murua, Maialen [1 ]
Santana, Roberto [2 ]
机构
[1] TECNALIA, Basque Res & Technol Alliance BRTA, Mikeletegi Pasealekua 7, San Sebastian 20009, Basque Country, Spain
[2] Univ Basque Country UPV EHU, Comp Sci & Artificial Intelligence Dept, Lardizabal Pasealekua 1, San Sebastian 20018, Basque Country, Spain
关键词
Flexible job-shop scheduling; Deep learning; Heterogeneous graph neural networks; Markov decision process; OPTIMIZATION;
D O I
10.1016/j.eswa.2024.125895
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Recent advancements in the flexible job-shop scheduling problem (FJSSP) are primarily based on deep reinforcement learning (DRL) due to its ability to generate high-quality, real-time solutions. However, many DRL approaches struggle with large action spaces, making the search for optimal solutions computationally intensive and impacting performance. Moreover, established techniques like exact methods and constraint programming (CP) often achieve better results for smaller instances. This paper aims to fully harness the strengths of these existing techniques by integrating CP within a deep learning (DL)-based methodology, leveraging the benefits of both. In this paper, we introduce a method that involves training a DL model using optimal solutions generated by CP, ensuring the model learns from high-quality data, thereby eliminating the need for the extensive exploration typical in DRL and enhancing overall performance. Further, we integrate CP into our DL framework to jointly construct solutions, utilizing DL for the initial complex stages and transitioning to CP for optimal resolution as the problem is simplified. Our hybrid approach has been extensively tested on three public FJSSP benchmarks, demonstrating superior performance over five state-of-the-art DRL approaches and a widely-used CP solver. Additionally, with the objective of exploring the application to other combinatorial optimization problems, promising preliminary results are presented on applying our hybrid approach to the traveling salesman problem, combining an exact method with a well-known DRL method.
引用
收藏
页数:15
相关论文
共 50 条
  • [1] Constraint programming: An efficient and practical approach to solving the job-shop problem
    Colombani, Yves
    Lecture Notes in Computer Science, 1118
  • [2] Machine learning and evolutionary optimization approach for solving the flexible job-shop scheduling problem
    Guo H.
    Yang J.
    Yang J.
    Journal of Intelligent and Fuzzy Systems, 2024, 46 (04): : 8845 - 8863
  • [3] A hybrid constraint programming local search approach to the job-shop scheduling problem
    Watson, Jean-Paul
    Beck, J. Christopher
    INTEGRATION OF AI AND OR TECHNIQUES IN CONSTRAINT PROGRAMMING FOR COMBINATORIAL OPTIMIZATION PROBLEMS, 2008, 5015 : 263 - +
  • [4] Solving the job-shop scheduling problem optimally by dynamic programming
    Gromicho, Joaquim A. S.
    van Hoorn, Jelke J.
    Saldanha-da-Gama, Francisco
    Timmer, Gerrit T.
    COMPUTERS & OPERATIONS RESEARCH, 2012, 39 (12) : 2968 - 2977
  • [5] Unified Genetic Algorithm Approach for Solving Flexible Job-Shop Scheduling Problem
    Park, Jin-Sung
    Ng, Huey-Yuen
    Chua, Tay-Jin
    Ng, Yen-Ting
    Kim, Jun-Woo
    APPLIED SCIENCES-BASEL, 2021, 11 (14):
  • [6] A distributed approach solving partially flexible job-shop scheduling problem with a Q-learning effect
    Bouazza, W.
    Sallez, Y.
    Beldjilali, B.
    IFAC PAPERSONLINE, 2017, 50 (01): : 15890 - 15895
  • [7] Constraint program for solving the job-shop problem
    Lect Notes Comput Sci, (510):
  • [8] A novel method for solving dynamic flexible job-shop scheduling problem via DIFFormer and deep reinforcement learning
    Wan, Lanjun
    Cui, Xueyan
    Zhao, Haoxin
    Fu, Long
    Li, Changyun
    COMPUTERS & INDUSTRIAL ENGINEERING, 2024, 198
  • [9] A new neighborhood structure for solving the flexible job-shop scheduling problem
    Huang X.
    Chen S.
    Zhou T.
    Sun Y.
    Xitong Gongcheng Lilun yu Shijian/System Engineering Theory and Practice, 2021, 41 (09): : 2367 - 2378
  • [10] A novel initialization method for solving flexible job-shop scheduling problem
    Shi Yang
    Zhang Guohui
    Gao Liang
    Yuan Kun
    CIE: 2009 INTERNATIONAL CONFERENCE ON COMPUTERS AND INDUSTRIAL ENGINEERING, VOLS 1-3, 2009, : 68 - +