Learning-based two-phase cooperative optimizer for distributed machine scheduling with heterogeneous factories and order priorities DOI Creative Commons
Tianyong Wu, Cong Luo, Youkou Dong

et al.

Egyptian Informatics Journal, Journal Year: 2023, Volume and Issue: 25, P. 100424 - 100424

Published: Dec. 8, 2023

In the realm of customized manufacturing, production cycles are often compressed to capture market opportunities swiftly. The blanking system stands as inaugural and pivotal phase in large equipment manufacturing. This study abstracts a novel problem from real-world systems, distributed unrelated parallel machine scheduling with heterogeneous factories order priorities (DUPMS-HP). presented work formulates bi-objective DUPMS-HP, aiming minimize both total weighted tardiness workload gap each machine. A learning-based two-phase cooperative optimizer (LCTPO) is introduced address this NP-hard problem, featuring: i) evolutionary algorithm during first stage for global search ensure diversity; ii) incorporation five problem-specific local strategies balance priority due date constraints. Additionally, reinforcement learning applied learn select best neighborhood operator elite solution, further enhancing diversity. effectiveness proposed validated through comparative analysis state-of-the-art algorithms on 20 instances. Experimental results affirm that LCTPO more adept at solving DUPMS-HP compared alternative algorithms.

Language: Английский

Surprisingly Popular-Based Adaptive Memetic Algorithm for Energy-Efficient Distributed Flexible Job Shop Scheduling DOI
Rui Li, Wenyin Gong, Ling Wang

et al.

IEEE Transactions on Cybernetics, Journal Year: 2023, Volume and Issue: 53(12), P. 8013 - 8023

Published: June 8, 2023

With the development of economy, distributed manufacturing has gradually become mainstream production mode. This work aims to solve energy-efficient flexible job shop scheduling problem (EDFJSP) while simultaneously minimizing makespan and energy consumption. Some gaps are stated following: 1) previous works usually adopt memetic algorithm (MA) with variable neighborhood search. However, local search (LS) operators inefficient due strong randomness; 2) confidence-based adaptive operator selection model follows experiences major crowds, which ignores efficient low weight, so it can not select really operator; 3) lack strategy save energy; 4) framework adopts LS all solutions, causes population converge too quickly diversity is extremely reduced. Thus, we propose a surprisingly popular-based MA (SPAMA) overcome above deficiencies. The contributions as follows: four problem-based employed improve convergence; popular degree (SPD) feedback-based self-modifying proposed find weight correct crowd decision making; full active decoding presented reduce consumption; an elite designed balance resources between global LS. In order evaluate effectiveness SPAMA, compared state-of-the-art algorithms on Mk DP benchmarks. results demonstrate superiority SPAMA state-of-art for solving EDFJSP.

Language: Английский

Citations

60
Deep reinforcement learning-based memetic algorithm for energy-aware flexible job shop scheduling with multi-AGV DOI

Fayong Zhang,

Rui Li, Wenyin Gong

et al.

Computers & Industrial Engineering, Journal Year: 2024, Volume and Issue: 189, P. 109917 - 109917

Published: Feb. 2, 2024

Language: Английский

Citations

26
Evolutionary computation and reinforcement learning integrated algorithm for distributed heterogeneous flowshop scheduling DOI
Rui Li, Ling Wang, Wenyin Gong

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2024, Volume and Issue: 135, P. 108775 - 108775

Published: June 12, 2024

Language: Английский

Citations

9
Improved Jaya algorithm for energy-efficient distributed heterogeneous permutation flow shop scheduling DOI
Qiwen Zhang, Zhen Tian

The Journal of Supercomputing, Journal Year: 2025, Volume and Issue: 81(2)

Published: Jan. 24, 2025

Language: Английский

Citations

1
Double DQN-Based Coevolution for Green Distributed Heterogeneous Hybrid Flowshop Scheduling With Multiple Priorities of Jobs DOI
Rui Li, Wenyin Gong, Ling Wang

et al.

IEEE Transactions on Automation Science and Engineering, Journal Year: 2023, Volume and Issue: 21(4), P. 6550 - 6562

Published: Nov. 1, 2023

Distributed manufacturing involving heterogeneous factories presents significant challenges to enterprises. Furthermore, the need prioritize various jobs based on order urgency and customer importance further complicates scheduling process. Consequently, this study addresses practical issue by tackling distributed hybrid flow shop problem with multiple priorities of (DHHFSP-MPJ). The primary objective is simultaneously minimize total weighted tardiness energy consumption. To solve DHHFSP-MPJ, a double deep Q-network-based co-evolution (D2QCE) developed four features: i) global local searches are allocated into two populations balance computational resources; ii) A heuristic strategy proposed obtain an initialized population great convergence diversity; iii) Four knowledge-based neighborhood structures accelerate converging. Next, Q-Network applied learn operator selection; iv) An energy-efficient presented save energy. verify effectiveness D2QCE, five state-of-the-art algorithms compared 20 instances real-world case. results numerical experiments indicate that: D2QN can fast only consuming few computation resources select best operator. Combining vastly improve performance evolutionary for solving scheduling. D2QCE has better than state-of-the-arts DHHFSP-MPJ Note Practitioners —This paper inspired encountered in blanking workshop systems within large engineering equipment. In scenario, come varying distinct due dates. Balancing these priority date constraints while efficiently considerable volume enhance enterprise profitability poses challenge. Thus, abstracted jobs. objectives minimizing delay Notably, model never been studied before. address this, we've formulated mixed-integer linear programming novel co-evolutionary algorithm Q-networks (DQN). Our approach introduces several key components. First, we present framework strike between search aspects. Additionally, devised three problem-specific enhancement strategies expedite convergence, which include initialization, techniques, energy-saving measures. learning process selecting optimal minimal resources, employ DQN. Experimental demonstrate superior our approach, outperforming when summary, work proposes extended DHHFSP provides case designing learning-assisted algorithm. However, online reinforcement (DRL) consumes additional time, generalization DRL needs be improved. future research, will consider dynamic events such as new insert change workshop. Moreover, end-to-end considered realize sustainable DRL.

Language: Английский

Citations

19
Minimizing tardiness and makespan for distributed heterogeneous unrelated parallel machine scheduling by knowledge and Pareto-based memetic algorithm DOI Creative Commons

Hua Wang,

Rui Li, Wenyin Gong

et al.

Egyptian Informatics Journal, Journal Year: 2023, Volume and Issue: 24(3), P. 100383 - 100383

Published: May 29, 2023

This work aims to deal with the distributed heterogeneous unrelated parallel machine scheduling problem (DHUPMSP) minimizing total tardiness (TDD) and makespan. To solve this complex combinatorial optimization problem, proposed a knowledge Pareto-based memetic algorithm (KPMA) which contains following features: 1) four heuristic rules are designed including shortest processing time rule, minimum factory workload finish earliest due date rule. Meanwhile, hybrid initialization is developed construct population great convergence diversity; 2) feature-based neighborhood structures increase success rate of local search; 3) simple elite strategy enhance usage historical solutions. Finally, evaluate performance KMPA, it compared five state-of-art run on 20 instances different scales. The results numerical experiments show that can efficiently save computation resources improve initialized convergence. In addition, knowledge-based vastly accelerate exploration. Moreover, diversity final non-dominated solutions set. KPMA has better than strong ability DHUPMSP.

Language: Английский

Citations

18
Reinforcement learning for distributed hybrid flowshop scheduling problem with variable task splitting towards mass personalized manufacturing DOI
Xin Chen, Yibing Li, Kaipu Wang

et al.

Journal of Manufacturing Systems, Journal Year: 2024, Volume and Issue: 76, P. 188 - 206

Published: Aug. 3, 2024

Language: Английский

Citations

8
An enhanced memetic algorithm with hierarchical heuristic neighborhood search for type-2 green fuzzy flexible job shop scheduling DOI
Kanglin Huang, Wenyin Gong, Chao Lu

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2023, Volume and Issue: 130, P. 107762 - 107762

Published: Dec. 26, 2023

Language: Английский

Citations

14
Research on computing task scheduling method for distributed heterogeneous parallel systems DOI Creative Commons

Xianzhi Cao,

Chong Chen, Shiwei Li

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 15, 2025

Abstract With the explosive growth of terminal devices, scheduling massive parallel task streams has become a core challenge for distributed platforms. For computing resource providers, enhancing reliability, shortening response times, and reducing costs are significant challenges, particularly in achieving energy efficiency through to realize green computing. This paper investigates heterogeneous flow problem minimize system consumption under time constraints. First, set independent tasks capable computation on terminals, is performed according computational capabilities each terminal. The modeled as mixed-integer nonlinear programming using Directed Acyclic Graph input model. Then, dynamic method based heuristic reinforcement learning algorithms proposed schedule flows. Furthermore, redundancy applied certain reliability analysis enhance fault tolerance improve service quality. Experimental results show that our can achieve improvements, by 14.3% compared existing approaches two practical workflow instances.

Language: Английский

Citations

0
Order acceptance and scheduling under eligibility, availability, and budget constraints in distributed heterogeneous flow shop production DOI
Fuli Xiong,

Lin Jing,

Chun Xiang

et al.

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127586 - 127586

Published: April 1, 2025

Language: Английский

Citations

0