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

Computers & Industrial Engineering, Journal Year: 2025, Volume and Issue: unknown, P. 110855 - 110855

Published: Jan. 1, 2025

Language: Английский

---

Manufacturing resource-based self-organizing scheduling using multi-agent system and deep reinforcement learning

Journal of Manufacturing Systems, Journal Year: 2025, Volume and Issue: 79, P. 179 - 198

Published: Jan. 24, 2025

Language: Английский

---

Multi-agent deep reinforcement learning based Predictive Maintenance on parallel machines

Robotics and Computer-Integrated Manufacturing, Journal Year: 2022, Volume and Issue: 78, P. 102406 - 102406

Published: July 8, 2022

In the context of Industry 4.0, companies understand advantages performing Predictive Maintenance (PdM). However, when moving towards PdM, several considerations must be carefully examined. First, they need to have a sufficient number production machines and relative fault data generate maintenance predictions. Second, adopt right approach, which, ideally, should self-adapt machinery, priorities organization, technician skills, but also able deal with uncertainty. Reinforcement learning (RL) is envisioned as key technique in this regard due its inherent ability learn by interacting through trials errors, very few RL-based frameworks been proposed so far literature, or are limited respects. This paper proposes new multi-agent approach that learns policy performed technicians, under uncertainty multiple machine failures. comprises RL agents partially observe state each coordinate decision-making scheduling, resulting dynamic assignment tasks technicians (with different skills) over set machines. Experimental evaluation shows our outperforms traditional policies (incl., corrective preventive ones) terms failure prevention downtime, improving ≈75% overall performance.

Language: Английский

---

Real-time scheduling for distributed permutation flowshops with dynamic job arrivals using deep reinforcement learning

Advanced Engineering Informatics, Journal Year: 2022, Volume and Issue: 54, P. 101776 - 101776

Published: Oct. 1, 2022

Language: Английский

---

Edge computing-based real-time scheduling for digital twin flexible job shop with variable time window

Robotics and Computer-Integrated Manufacturing, Journal Year: 2022, Volume and Issue: 79, P. 102435 - 102435

Published: July 31, 2022

Production scheduling is the central link between enterprise production and operation management also key to realising efficient, high-quality sustainable production. However, in real-world manufacturing, frequent occurrence of abnormal disturbance leads deviation scheduling, which affects accuracy reliability execution. The traditional dynamic methods (TDSMs) cannot solve this problem effectively. This paper presents a real-time digital twin flexible job shop (R-DTFJSS) method with edge computing address issue. Firstly, an overall framework R-DTFJSS proposed realise (RS) through interaction physical workshop (PW) virtual (VW). Secondly, implementation process designed allocation. Then, obtain optimal RS result, improved Hungarian algorithm (IHA) adopted. Finally, case simulation from industrial cooperative described analysed verify effectiveness method. results show that compared TDSMs, can effectively deal unexpected disturbances process.

Language: Английский

---

A Pareto-based two-stage evolutionary algorithm for flexible job shop scheduling problem with worker cooperation flexibility

Robotics and Computer-Integrated Manufacturing, Journal Year: 2023, Volume and Issue: 82, P. 102534 - 102534

Published: Jan. 30, 2023

Language: Английский

---

Dynamic scheduling for flexible job shop with insufficient transportation resources via graph neural network and deep reinforcement learning

Computers & Industrial Engineering, Journal Year: 2023, Volume and Issue: 186, P. 109718 - 109718

Published: Oct. 31, 2023

Language: Английский

---

Dynamic production scheduling towards self-organizing mass personalization: A multi-agent dueling deep reinforcement learning approach

Journal of Manufacturing Systems, Journal Year: 2023, Volume and Issue: 68, P. 242 - 257

Published: April 9, 2023

Language: Английский

---

An Improved Genetic Algorithm for Solving the Multi-AGV Flexible Job Shop Scheduling Problem

Sensors, Journal Year: 2023, Volume and Issue: 23(8), P. 3815 - 3815

Published: April 7, 2023

In real manufacturing environments, the number of automatic guided vehicles (AGV) is limited. Therefore, scheduling problem that considers a limited AGVs much nearer to production and very important. this paper, we studied flexible job shop with (FJSP-AGV) propose an improved genetic algorithm (IGA) minimize makespan. Compared classical algorithm, population diversity check method was specifically designed in IGA. To evaluate effectiveness efficiency IGA, it compared state-of-the-art algorithms for solving five sets benchmark instances. Experimental results show proposed IGA outperforms algorithms. More importantly, current best solutions 34 instances four data were updated.

Language: Английский

---

A DRL-Based Reactive Scheduling Policy for Flexible Job Shops With Random Job Arrivals

IEEE Transactions on Automation Science and Engineering, Journal Year: 2023, Volume and Issue: 21(3), P. 2912 - 2923

Published: May 12, 2023

In real-life production systems, arrivals of jobs are usually unpredictable, which makes it necessary to develop solid reactive scheduling policies meet delivery requirements. Deep reinforcement learning (DRL) based methods capable quickly responding dynamic events by from the training data. However, most policy networks in DRL algorithms trained choose priority dispatching rules (PDR), thus, some extent, efficiency obtained plans is limited performance PDRs. This paper investigates a flexible job shop problem with random for total tardiness minimization. A DRL-based method, proximal optimization attention-based network (PPO-APN), proposed make real-time decisions environment, where (APN) able directly select pending distinguished action space that consists Additionally, global/local reward function (GLRF) designed address sparsity issue during processes. The PPO-APN tested on randomly generated instances different configurations, and compared frequently-used PDRs methods. Numerical experimental results indicate APN GLRF components significantly improve efficiency, shows better overall other Note Practitioners —This work motivated typical scenario discrete manufacturing orders arrive at floor require be scheduled short time ensure on-time delivery. Previous research tends apply suitable ease implementation. Nevertheless, can selected rather limited, thus many possible high-quality ignored. first sorts all unscheduled heuristic algorithm, puts top-ranked pool. When machine becomes available, will pool as next processing task. selection represented novel network, powerful algorithm. aforementioned process repeatedly executed simulation environment collect Therefore, after being certain period time, become smarter applied right real-time. method has been proved more efficient than approaches, effective wide variety scenarios, such automobile electronics industries. Moreover, further extended problems characteristics via adding constraints or re-defining calculations completion operations accordingly.

Language: Английский

---