HGNP: A PCA-based heterogeneous graph neural network for a family distributed flexible job shop

Computers & Industrial Engineering, Год журнала: 2025, Номер unknown, С. 110855 - 110855

Опубликована: Янв. 1, 2025

Язык: Английский

---

Dynamic Job-Shop Scheduling via Graph Attention Networks and Deep Reinforcement Learning

IEEE Transactions on Industrial Informatics, Год журнала: 2024, Номер 20(6), С. 8662 - 8672

Опубликована: Март 21, 2024

The dynamic job-shop scheduling problem (DJSSP) is an advanced form of the classical (JSSP), incorporating events that make it even more challenging. This article proposes a novel approach involving deep reinforcement learning and graph neural networks to solve this optimization problem. To effectively model DJSSP, we use disjunctive graph, designing specific node features reflect unique characteristics JSSP with machine breakdowns stochastic job arrivals. Our proposed method can dynamically adapt occurrence disruptions, ensuring accurately reflects current state environment. Furthermore, attention mechanism prioritize crucial nodes while discarding irrelevant ones. study applies learn embeddings, serving as input for actor–critic model. proximal policy then utilized train model, which assists in operations machines. We conducted extensive experiments static public environments. Experimental results indicate our superior state-of-the-art methods.

Язык: Английский

---

Fifty years of research in scheduling – theory and applications

European Journal of Operational Research, Год журнала: 2025, Номер unknown

Опубликована: Фев. 1, 2025

Язык: Английский

---

Multi-Agent Reinforcement Learning for Job Shop Scheduling in Dynamic Environments

Sustainability, Год журнала: 2024, Номер 16(8), С. 3234 - 3234

Опубликована: Апрель 12, 2024

In response to the challenges of dynamic adaptability, real-time interactivity, and optimization posed by application existing deep reinforcement learning algorithms in solving complex scheduling problems, this study proposes a novel approach using graph neural networks complete task job shop scheduling. A distributed multi-agent architecture (DMASA) is constructed maximize global rewards, modeling intelligent manufacturing problem as sequential decision represented graphs Graph Embedding–Heterogeneous Neural Network (GE-HetGNN) encode state nodes map them optimal strategy, including machine matching process selection strategies. Finally, an actor–critic architecture-based proximal policy algorithm employed train network optimize decision-making process. Experimental results demonstrate that proposed framework exhibits generalizability, outperforms commonly used rules RL-based methods on benchmarks, shows better stability than single-agent architectures, breaks through instance-size constraint, making it suitable for large-scale problems. We verified feasibility our method specific experimental environment. The research can achieve formal mapping with physical processing workshops, which aligns more closely real-world green issues makes easier subsequent researchers integrate actual environments.

Язык: Английский

---

Learning to Dispatch for Flexible Job Shop Scheduling Based on Deep Reinforcement Learning via Graph Gated Channel Transformation

IEEE Access, Год журнала: 2024, Номер 12, С. 50935 - 50948

Опубликована: Янв. 1, 2024

In addressing the Flexible Job Shop Scheduling Problem (FJSP), deep reinforcement learning eliminates need for mathematical modeling of problem, requiring only interaction with real environment to learn effective strategies. Using disjunctive graphs as state representation has proven be a particularly method. Additionally, attention mechanisms enable rapid focus on relevant features. However, due unique structure mechanisms, current methods fail provide strategies after changes in scale. To resolve this issue, we propose an end-to-end framework FJSP. Initially, introduce lightweight model, Graph Gated Channel Transformation (GGCT), identify characteristics workpieces being scheduled at decision-making moment, while suppressing redundant Subsequently, address inability scale, modify expression graph features, channeling global features into different channels capture information moment effectively. Comparative analysis generated and classical datasets shows our model reduces average makespan significantly, from 8.243% 7.037% 10.08% 8.69%, respectively.

Язык: Английский

---

Graph neural networks for job shop scheduling problems: A survey

Computers & Operations Research, Год журнала: 2024, Номер unknown, С. 106914 - 106914

Опубликована: Ноя. 1, 2024

Язык: Английский

---

Low-Carbon Flexible Job Shop Scheduling Problem Based on Deep Reinforcement Learning

Sustainability, Год журнала: 2024, Номер 16(11), С. 4544 - 4544

Опубликована: Май 27, 2024

As the focus on environmental sustainability sharpens, significance of low-carbon manufacturing and energy conservation continues to rise. While traditional flexible job shop scheduling strategies are primarily concerned with minimizing completion times, they often overlook consumption machines. To address this gap, paper introduces a novel solution utilizing deep reinforcement learning. The study begins by defining Low-carbon Flexible Job Shop Scheduling problem (LC-FJSP) constructing disjunctive graph model. A sophisticated representation, based Markov Decision Process (MDP), incorporates attention network featuring multi-head modules pooling techniques, aimed at boosting model’s generalization capabilities. Additionally, Bayesian optimization is employed enhance refinement process, method benchmarked against conventional models. empirical results indicate that our algorithm markedly enhances efficiency 5% 12% reduces carbon emissions 3% 8%. This work not only contributes new insights methods realm green production but also underscores its considerable theoretical practical implications.

Язык: Английский

---

A survey on vehicle–drone cooperative delivery operations optimization: Models, methods, and future research directions

Swarm and Evolutionary Computation, Год журнала: 2024, Номер 92, С. 101780 - 101780

Опубликована: Ноя. 26, 2024

Язык: Английский

---

Learning to schedule dynamic distributed reconfigurable workshops using expected deep Q-network

Advanced Engineering Informatics, Год журнала: 2023, Номер 59, С. 102307 - 102307

Опубликована: Дек. 15, 2023

Язык: Английский

---

Deep reinforcement learning driven trajectory-based meta-heuristic for distributed heterogeneous flexible job shop scheduling problem

Swarm and Evolutionary Computation, Год журнала: 2024, Номер 91, С. 101753 - 101753

Опубликована: Окт. 9, 2024

Язык: Английский

---