Heuristic deep reinforcement learning for online packing

Chengbo Yang; Yonggui Lü; Jing Du; Ting Liu; Ronggui Dao

doi:10.1117/12.3011479

7 December 2023 Heuristic deep reinforcement learning for online packing

Chengbo Yang, Yonggui Lü, Jing Du, Ting Liu, Ronggui Dao

Proceedings Volume 12941, International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 2023); 1294158 (2023) https://doi.org/10.1117/12.3011479
Event: Third International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 203), 2023, Yinchuan, China

Abstract

In order to improve the utilization rate of loading space, a model of online packing problem was established. In this paper, we introduce an advanced technique for incorporating heuristic knowledge distillation into reinforcement learning agents. Our approach utilizes a two-stage learning process, where a teacher model, trained with heuristics, guides a student model in acquiring knowledge directly from the environment through reinforcement learning. To assess the performance of our method, we applied it to a logistics loading scenario, addressing an online packing problem in simulated world environments. Experimental results revealed that our method outperforms both the heuristic and conventional RL models, highlighting the effectiveness of the heuristic knowledge distillation architecture in enhancing RL agent performance across diverse settings.

Citation Download Citation

Chengbo Yang, Yonggui Lü, Jing Du, Ting Liu, and Ronggui Dao "Heuristic deep reinforcement learning for online packing", Proc. SPIE 12941, International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 2023), 1294158 (7 December 2023); https://doi.org/10.1117/12.3011479

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available