Research on a Dynamic Algorithm of Simulated Warehouse Environment Based on Q-learning
DOI:
https://doi.org/10.54097/2qwk1s93Keywords:
Q-learning; Path planning; intelligent storage; Machine learning.Abstract
This paper uses Q-learning and warehouse simulation environment path planning algorithm training and research. The application environment is a dynamic and complex environment for a real warehouse. However, traditional research is more focused on exploring the Q-learning algorithm itself, focusing on the algorithm itself rather than the application, which lacks practicability. Therefore, this study will take the real map of the warehouse as a sample can combine Q-learning to simulate the algorithm closer to reality. The algorithm simulation will be carried out in a unique three-stage form to fit the influence of dynamic and changeable on the algorithm training in the actual situation. After algorithm training and simulation, the visual path result of Q-learning in the warehouse map is obtained. Based on the result, the advantages and disadvantages of Q-learning in warehouse path planning are discussed, and suggestions for future algorithm improvement. Q-learning algorithm is a promising algorithm. Through its research, more mature path planning results can be obtained, and it can be used to optimize cargo transportation efficiency and warehouse management mode in industrial production, logistics, and other fields, to obtain actual benefits.
Downloads
References
[1] Li, W., & Smith, J. Warehouse management systems and logistics optimization. International Journal of Logistics Management, 2017, 28(2), 23-39.
[2] Sutton, R. S., & Barto, A. G. Reinforcement Learning: An Introduction. MIT Press, 2018.
[3] Wang, X., & Lu, J. Multi-agent Q-learning approach for warehouse task coordination. IEEE Transactions on Automation Science and Engineering, 2019, 16(3), 728-736.
[4] Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., et al. Human-level control through deep reinforcement learning. Nature, 2015, 518(7540), 529-533.
[5] Li, X., & Chen, Y. Dynamic Q-learning algorithm for uncertain warehouse environments. IEEE Access, 2020, 8, 75235-75245.
[6] Van Hasselt, H., Guez, A., & Silver, D. Deep reinforcement learning with double Q-learning. AAAI Conference on Artificial Intelligence, 2016.
[7] Chen, G., & Zhang, H. Reinforcement learning challenges in dynamic environments: A survey. ACM Computing Surveys, 2021, 53(4), 1-35.
[8] Kumar, A., & Verma, R. Task scheduling optimization using reinforcement learning in large-scale warehouse systems. International Journal of Advanced Robotic Systems, 2021, 18(1), 22-31.
[9] Lemos M R, de Souza A V R, de Lira R S, et al. Robot training and navigation through the deep Q-Learning algorithm. 2021 IEEE International Conference on Consumer Electronics (ICCE). IEEE, 2021, 1-6.
[10] Jiang L, Huang H, Ding Z. Path planning for intelligent robots based on deep Q-learning with experience replay and heuristic knowledge. IEEE/CAA Journal of Automatica Sinica, 2019, 7(4): 1179-1189.
[11] Cestero J, Quartulli M, Metelli A M, et al. Storehouse: A reinforcement learning environment for optimizing warehouse management. 2022 International Joint Conference on Neural Networks (IJCNN). IEEE, 2022, 1-9.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Highlights in Science, Engineering and Technology
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
