.png){kind=icon}
.png){kind=icon}
Return
An efficient cloud manufacturing service composition approach using deep reinforcement learning
.png){kind=icon}
PRE
AI
DOI：10.1016/j.cie.2024.110446.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Cloud Manufacturing (CMfg) revolutionizes manufacturing by providing resources as cloud-based services. The main challenge in CMfg is identifying the best combination of these services to meet customer needs efficiently. This paper explores the potential of Deep Reinforcement Learning (DRL), a technology with successful applications in fields such as healthcare and transportation, to tackle scheduling challenges in CMfg. A tailored DRL environment for CMfg is introduced, along with a novel DRL-based algorithm designed to optimize service composition in CMfg. Comparative evaluations against existing algorithms, both in the new DRL-based CMfg environment and a benchmark DRL environment, consistently showed the superior performance of the proposed algorithm. Additionally, the study analyzes different combinations of weight coefficients and probabilities of service failure within the new CMfg environment, revealing their significant impact on scheduling outcomes. These findings underscore both the robustness and efficiency of the proposed solution, showcasing DRL's potential to significantly enhance CMfg operations.
Keywords:
Cloud manufacturing
Service composition
Deep reinforcement learning
Scheduling
Optimization
Journal
IF:
6.5
Papers:
1.0W
Citations:
3.8W