Deep reinforcement learning in smart manufacturing: A review and prospects

Li, Chengxi; Zheng, Pai; Yin, Yue; Wang, Baicun; Wang, Lihui

doi:10.1016/j.cirpj.2022.11.003

Cited by 125 publications

(28 citation statements)

References 224 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In recent years, RL applications sprung up across the manufacturing field with exponential publication growth year by year as Li et al [14] stated. The authors analyzed 264 different publications between 2013 and October 2022 and found, that optimizing energy consumption as well as costs and reducing reliance on expert knowledge as the main objectives of these applications.…”

Section: Reinforcement Learning For Process Optimizationmentioning

confidence: 99%

Advanced automatic pass schedule design for hot rolling by coupling reinforcement learning with a fast rolling model

Idzik

2023

Materials Research Proceedings

View full text Add to dashboard Cite

Abstract. Rolling is a well-established forming process for producing finished or semi-finished products in various industries. Although highly automated, most rolling processes are designed manually by experts based on their knowledge, highly specialized heuristics and analytical process models or numerical simulations. This manual design approach does not lead to an optimization accounting for multiple objectives. Previous work [1] has shown the potential of coupling reinforcement learning (RL) with fast analytical rolling models (FRM) to optimize hot rolling processes. However, the designed pass schedules do not robustly reach the desired final height within typical industrial tolerances. Therefore, in this paper the existing approach of coupling RL with an FRM is extended by dynamically ranges for height reductions. This extension guarantees that the target height is always reached exactly. In addition to the height reduction, the RL algorithm can determine the inter-pass time, initial slab temperature and rolling velocity. For the optimization, an objective function, called reward function, considering all relevant optimization objectives such as the final grain size and energy consumption, was developed. An exemplary training was performed for a defined starting (140 mm) and final height (25 mm). The resulting, automatically designed pass schedules reach the target height and fulfill all defined optimization objective including the required average austenite grain size.

show abstract

Section: Reinforcement Learning For Process Optimizationmentioning

confidence: 99%

Advanced automatic pass schedule design for hot rolling by coupling reinforcement learning with a fast rolling model

Idzik

2023

Materials Research Proceedings

View full text Add to dashboard Cite

show abstract

“…Among many methods, the reinforcement learning algorithm is more suitable [36][37][38]. It can allow the ship agent to interact with the environment in real-time and improve the collision avoidance ability by constantly trying to accumulate experience, which is very similar to the experience accumulation process of people.…”

Section: The Proposed Deep Q Network(dqn)mentioning

confidence: 99%

Improved DQN for Dynamic Obstacle Avoidance and Ship Path Planning

Yang

Han

2023

Algorithms

View full text Add to dashboard Cite

The avoidance of collisions among ships requires addressing various factors such as perception, decision-making, and control. These factors pose many challenges for autonomous collision avoidance. Traditional collision avoidance methods have encountered significant difficulties when used in autonomous collision avoidance. They are challenged to cope with the changing environment and harsh motion constraints. In the actual navigation of ships, it is necessary to carry out decision-making and control under the constraints of ship manipulation and risk. From the implementation process perspective, it is a typical sequential anthropomorphic decision-making problem. In order to solve the sequential decision problem, this paper improves DQN by setting a priority for sample collection and adopting non-uniform sampling, and it is applied to realize the intelligent collision avoidance of ships. It also verifies the performance of the algorithm in the simulation environment.

show abstract

“…Driven by the need of customizable device design and the requirement of sustainability, future fiber-based biofabrication would develop toward on demand production while minimizing the environmental footprints . Artificial intelligence could provide efficient device structural and functionality designs to maximize higher customization freedom and supply chain robustness. , Material- and power-light fiber printing technologies promise on demand and on site/in situ fiber fabrication (i.e., with a movable robotic printing platform commended by smartphones). Looking ahead, we can expect the emergence of multifunctional, environmentally friendly, and on demand fiber devices to revolutionize bioelectronic technologies for fundamental neuroscience research, 3D cell culture and microphysiological systems, and wearable sensors and “Fiber-of-Things” (FoT).…”

Section: Conclusion and Outlookmentioning

confidence: 99%

Biointerface Fiber Technology from Electrospinning to Inflight Printing

Wang,

Ka,

Pan

et al. 2023

ACS Appl. Mater. Interfaces

View full text Add to dashboard Cite

Building two-dimensional (2D) and three-dimensional (3D) micro-and nanofibril structures with designable patterns and functionalities will offer exciting prospects for numerous applications spanning from permeable bioelectronics to tissue engineering scaffolds. This Spotlight on Applications highlights recent technological advances in fiber printing and patterning with functional materials for biointerfacing applications. We first introduce the current state of development of micro-and nanofibers with applications in biology and medical wearables. We then describe our contributions in developing a series of fiber printing techniques that enable the patterning of functional fiber architectures in three dimensions. These fiber printing techniques expand the material library and device designs, which underpin technological capabilities from enabling fundamental studies in cell migration to customizable and ecofriendly fabrication of sensors. Finally, we provide an outlook on the strategic pathways for developing the next-generation bioelectronics and "Fiber-of-Things" (FoT) using nano/micro-fibers as architectural building blocks.

show abstract

Deep reinforcement learning in smart manufacturing: A review and prospects

Cited by 125 publications

References 224 publications

Advanced automatic pass schedule design for hot rolling by coupling reinforcement learning with a fast rolling model

Advanced automatic pass schedule design for hot rolling by coupling reinforcement learning with a fast rolling model

Improved DQN for Dynamic Obstacle Avoidance and Ship Path Planning

Biointerface Fiber Technology from Electrospinning to Inflight Printing

Contact Info

Product

Resources

About