Gantry Scheduling for Two-Machine One-Buffer Composite Work Cell by Reinforcement Learning

2017 ◽  
Author(s):  
Jorge Arinez ◽  
Xinyan Ou ◽  
Qing Chang
Keyword(s):  
Reinforcement Learning ◽  
Learning Algorithm ◽  
Production Lines ◽  
Serial Production ◽  
Q Learning ◽  
Proper Actions ◽  
Work Cell ◽  
Scheduling Policy ◽  
Serial Production Lines

In this paper, a manufacturing work cell with a gantry that is in charge of moving materials/parts between machines and buffers is considered. With the effect of the gantry movement, the system performance becomes quite different from traditional serial production lines. In this paper, reinforcement learning is used to develop a gantry scheduling policy in order to improve system production. The gantry learns to take proper actions under different situations to reduce system production loss by using Q-Learning algorithm and finds the optimal moving policy. A two-machine one-buffer work cell with a gantry is used for case study, by which reinforcement learning is applied. Compare with the FCFS policy, the fidelity and effectiveness of the reinforcement learning method are also demonstrated.

scholarly journals Aircraft Maintenance Check Scheduling Using Reinforcement Learning

Aerospace ◽  
2021 ◽  
Vol 8 (4) ◽  
pp. 113
Author(s):  
Pedro Andrade ◽  
Catarina Silva ◽  
Bernardete Ribeiro ◽  
Bruno F. Santos
Keyword(s):  
Reinforcement Learning ◽  
Time Horizon ◽  
Learning Algorithm ◽  
Initial Conditions ◽  
Q Learning ◽  
Scheduling Policy ◽  
Real Scenario ◽  
Maintenance Plan ◽  
Small Disturbances

This paper presents a Reinforcement Learning (RL) approach to optimize the long-term scheduling of maintenance for an aircraft fleet. The problem considers fleet status, maintenance capacity, and other maintenance constraints to schedule hangar checks for a specified time horizon. The checks are scheduled within an interval, and the goal is to, schedule them as close as possible to their due date. In doing so, the number of checks is reduced, and the fleet availability increases. A Deep Q-learning algorithm is used to optimize the scheduling policy. The model is validated in a real scenario using maintenance data from 45 aircraft. The maintenance plan that is generated with our approach is compared with a previous study, which presented a Dynamic Programming (DP) based approach and airline estimations for the same period. The results show a reduction in the number of checks scheduled, which indicates the potential of RL in solving this problem. The adaptability of RL is also tested by introducing small disturbances in the initial conditions. After training the model with these simulated scenarios, the results show the robustness of the RL approach and its ability to generate efficient maintenance plans in only a few seconds.

scholarly journals c-Bottlenecks in serial production lines: Identification and application

2001 ◽  
Vol 7 (6) ◽  
pp. 543-578 ◽  
Author(s):  
S.-Y. Chiang ◽  
C.-T. Kuo ◽  
S. M. Meerkov
Keyword(s):  
Cycle Time ◽  
Production Line ◽  
Markovian Model ◽  
Production Lines ◽  
Serial Production ◽  
Automotive Engine ◽  
Machine Reliability ◽  
Bottleneck Identification ◽  
Serial Production Lines

The bottleneck of a production line is a machine that impedes the system performance in the strongest manner. In production lines with the so-called Markovian model of machine reliability, bottlenecks with respect to the downtime, uptime, and the cycle time of the machines can be introduced. The two former have been addressed in recent publications [1] and [2]. The latter is investigated in this paper. Specifically, using a novel aggregation procedure for performance analysis of production lines with Markovian machines having different cycle time, we develop a method for c-bottleneck identification and apply it in a case study to a camshaft production line at an automotive engine plant.

scholarly journals An Empirical Investigation of Transfer Effects for Reinforcement Learning

2020 ◽  
Vol 2020 ◽  
pp. 1-10
Author(s):  
Jung-Sing Jwo ◽  
Ching-Sheng Lin ◽  
Cheng-Hsiung Lee ◽  
Ya-Ching Lo
Keyword(s):  
Reinforcement Learning ◽  
Empirical Investigation ◽  
Learning Algorithm ◽  
Q Learning ◽  
Sorting Problem ◽  
Long Time ◽  
The Difference ◽  
Reinforcement Model ◽  
The Brain

Previous studies have shown that training a reinforcement model for the sorting problem takes very long time, even for small sets of data. To study whether transfer learning could improve the training process of reinforcement learning, we employ Q-learning as the base of the reinforcement learning algorithm, apply the sorting problem as a case study, and assess the performance from two aspects, the time expense and the brain capacity. We compare the total number of training steps between nontransfer and transfer methods to study the efficiencies and evaluate their differences in brain capacity (i.e., the percentage of the updated Q-values in the Q-table). According to our experimental results, the difference in the total number of training steps will become smaller when the size of the numbers to be sorted increases. Our results also show that the brain capacities of transfer and nontransfer reinforcement learning will be similar when they both reach a similar training level.

scholarly journals RFLMDA: A Novel Reinforcement Learning-Based Computational Model for Human MicroRNA-Disease Association Prediction

Biomolecules ◽  
2021 ◽  
Vol 11 (12) ◽  
pp. 1835
Author(s):  
Linqian Cui ◽  
You Lu ◽  
Jiacheng Sun ◽  
Qiming Fu ◽  
Xiao Xu ◽  
...  
Keyword(s):  
Reinforcement Learning ◽  
Learning Algorithm ◽  
Computational Simulation ◽  
Predictive Performance ◽  
Human Diseases ◽  
Gastric Neoplasms ◽  
Optimal Weight ◽  
Q Learning ◽  
Disease Associations

Numerous studies have confirmed that microRNAs play a crucial role in the research of complex human diseases. Identifying the relationship between miRNAs and diseases is important for improving the treatment of complex diseases. However, traditional biological experiments are not without restrictions. It is an urgent necessity for computational simulation to predict unknown miRNA-disease associations. In this work, we combine Q-learning algorithm of reinforcement learning to propose a RFLMDA model, three submodels CMF, NRLMF, and LapRLS are fused via Q-learning algorithm to obtain the optimal weight S. The performance of RFLMDA was evaluated through five-fold cross-validation and local validation. As a result, the optimal weight is obtained as S (0.1735, 0.2913, 0.5352), and the AUC is 0.9416. By comparing the experiments with other methods, it is proved that RFLMDA model has better performance. For better validate the predictive performance of RFLMDA, we use eight diseases for local verification and carry out case study on three common human diseases. Consequently, all the top 50 miRNAs related to Colorectal Neoplasms and Breast Neoplasms have been confirmed. Among the top 50 miRNAs related to Colon Neoplasms, Gastric Neoplasms, Pancreatic Neoplasms, Kidney Neoplasms, Esophageal Neoplasms, and Lymphoma, we confirm 47, 41, 49, 46, 46 and 48 miRNAs respectively.

Sign in / Sign up

Export Citation Format

Share Document