Deep reinforcement learning for continuous wood drying production line control

Article Computer Science, Artificial Intelligence

Responsive Mixed-initiative System for Reoptimization of Mixed-integer Programming

Marc-Andre Menard et al.

Summary: The mixed-initiative system allows users to quickly find new optimal solutions by predicting user behavior and precomputing solutions that meet preferences. Tested on a wood industry problem, it shows promising results for future use.

INTERNATIONAL JOURNAL ON ARTIFICIAL INTELLIGENCE TOOLS (2022)

Add to Collection

Article Computer Science, Interdisciplinary Applications

Reinforcement learning for industrial process control: A case study in flatness control in steel industry

Jifei Deng et al.

Summary: This paper proposes a novel approach that combines ensemble learning with reinforcement learning methods for strip rolling control. The multi-actor PPO method is introduced to improve the control effectiveness. Simulation results demonstrate that the proposed method outperforms conventional control methods and state-of-the-art reinforcement learning methods in terms of process capability and smoothness.

COMPUTERS IN INDUSTRY (2022)

Add to Collection

Article Engineering, Marine

Actor-Critic reinforcement learning for optimal design of piping support constraint combinations

Jong-Ho Ham et al.

Summary: Piping stress analysis and design are crucial for ensuring safety, and selecting the appropriate combination of constraints is a time-consuming and challenging task. This paper proposes a reinforcement learning strategy for determining the optimal combination of piping support constraints and demonstrates the suitability and effectiveness of the algorithm through experimental verification.

INTERNATIONAL JOURNAL OF NAVAL ARCHITECTURE AND OCEAN ENGINEERING (2022)

Add to Collection

Article Computer Science, Artificial Intelligence

Designing an adaptive production control system using reinforcement learning

Andreas Kuhnle et al.

Summary: Reinforcement learning shows potential advantages in modern production systems, especially when facing dynamic and complex environments with limited domain knowledge. Research has identified robust RL designs that can be applied to adaptive production control systems.

JOURNAL OF INTELLIGENT MANUFACTURING (2021)

Add to Collection

Article Computer Science, Interdisciplinary Applications

A deep reinforcement learning based multi-criteria decision support system for optimizing textile chemical process

Zhenglei He et al.

Summary: The paper proposes a decision support system that combines intelligent data based models of random forest (RF) and a human knowledge-based multi-criteria structure of analytical hierarchical process (AHP) for the textile manufacturing process. It also describes the textile chemical manufacturing process as a Markov decision process (MDP) and employs Deep Q networks (DQN) for optimization, showing effectiveness in a case study of optimizing a textile ozonation process.

COMPUTERS IN INDUSTRY (2021)

Add to Collection

Article Computer Science, Information Systems

Dynamic Charging Scheme Problem With Actor-Critic Reinforcement Learning

Meiyi Yang et al.

Summary: In this article, a novel dynamic charging scheme based on actor-critic reinforcement learning algorithm in WRSN is proposed. The use of GRUs to capture the relationships of charging actions in time sequence is introduced. Extensive simulations show that the proposed ACRL algorithm surpasses heuristic algorithms in average lifetime and tour length.

IEEE INTERNET OF THINGS JOURNAL (2021)

Add to Collection

Editorial Material Automation & Control Systems

Sim2Real in Robotics and Automation: Applications and Challenges

Sebastian Hofer et al.

Summary: Large-scale automation relies on computer simulation to design, validate, and continuously improve complex processes, as well as provide insight for future investments. Despite successful applications in industries such as circuit simulation, finite element methods, and computer-aided design, current simulators fall short in accurately modeling physical phenomena like friction, impact, and deformation.

IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING (2021)

Add to Collection

Article Computer Science, Interdisciplinary Applications

Planning for automatic product assembly using reinforcement learning

Heng Zhang et al.

Summary: This paper proposes a reinforcement learning method for automatic operation of mechanical product assembly, aiming to improve efficiency and accuracy by building machine learning models and developing searching strategies. The method has been applied in the assembly improvement of function modules of an industrial machine.

COMPUTERS IN INDUSTRY (2021)

Add to Collection

Article Computer Science, Interdisciplinary Applications