Education, Science, Technology, Innovation and Life
Open Access
Sign In

Optimization of Multi-Direction Material Supply Demand Based on Model Algorithm

Download as PDF

DOI: 10.23977/ieim.2024.070115 | Downloads: 9 | Views: 204

Author(s)

Jie Ai 1, Guiming Chen 1

Affiliation(s)

1 Xi'an Research Institute of High-Tech, Xi'an, China

Corresponding Author

Jie Ai

ABSTRACT

The content of this study is to realize the real-time optimization of supply chain and improve the overall efficiency of supply chain by comprehensively considering various algorithms or models and comprehensively considering the needs of all parties in the supply chain. Through comparative experiments, this paper verifies that the proposed algorithm based on deep reinforcement learning has significant advantages in meeting the demand of multi-direction material supply. This study designed a series of key experiments to comprehensively evaluate the Deep Q-network DQN (Deep Q-Leaning Network) algorithm, GA (Genetic Algorithm) and PSO (Particle Swarm Optimization) algorithm and Q-Learning (Quintuple Learning) on multi-direction material supply chain optimization. In the benchmark performance test, the DQN algorithm has the lowest total supply cost of $1,200, the algorithm evaluation meets the demand in 2 hours, and the satisfaction rate reaches 95%, which is far higher than other algorithms. It can be seen from the dynamic demand adaptability experiment that DQN algorithm is highly adaptable and flexible in responding to market demand fluctuations. The average response time of DQN algorithm evaluation is 2 hours in season, and the response time of DQN algorithm can also be maintained at 2.2 hours in the case of demand surge caused by emergencies. The above data proves its superior performance in dynamic environments. The robust robustness of the DQN algorithm was also further confirmed in the robustness and exception handling experiments, where DQN showed the shortest recovery time of 1.5 hours and the lowest cost impact of a 5% cost increase in the face of anomalies such as supply point failures, demand surges and transportation delays. From the above experimental data, it can be seen that DQN algorithm shows excellent benchmark performance, dynamic adaptability and robustness in the multi-direction material supply chain optimization problem, which can be said to be an effective and reliable solution. 

KEYWORDS

Material Supply Chain Management, DQN Algorithm, Optimization Algorithm, Deep Reinforcement Learning, Robustness

CITE THIS PAPER

Jie Ai, Guiming Chen, Optimization of Multi-Direction Material Supply Demand Based on Model Algorithm. Industrial Engineering and Innovation Management (2024) Vol. 7: 113-120. DOI: http://dx.doi.org/10.23977/ieim.2024.070115.

REFERENCES

[1] Rong Bo, Feng Aifen, Lou Xinxin. Research on the location of agricultural product supply points in Henan Province based on operation research optimization. Operation Research and Fuzzy Science, 2023, 13(3):2058-2066.
[2] Ma Bowen, Wei Yuguang, Fang Bo, et al. Research on Optimization of dynamic railway flow organization for Station and station integration. Journal of Railway Science, 2023, 45(5):1-11.
[3] Wu Di, Cheng Xu, Zhang Wangyuhui. Research on fresh cargo route optimization based on improved genetic algorithm. Value Engineering, 2023, 42(17):44-46.
[4] Zhang Yingting. Optimal Path planning model of multi-mode ship segmental logistics transportation. Ship Science and Technology, 2020, 42(08):179-181.
[5] Kuang Yujie, Zhao Jiahong. Dangerous goods storage and transportation site-route selection problem under continuous time-varying risks. China Safety Science Journal, 2022, 32(4):185-191.
[6] Wu Peng, Li Ze, Ji Haitao. Green multimodal transport path and speed optimization considering emission control area. Transportation Systems Engineering and Information, 2023, 23(3):20-29.
[7] Fakhrzad M B, Goodarzian F. A new multi-objective mathematical model for a Citrus supply chain network design: Metaheuristic algorithms. Journal of Optimization in Industrial Engineering, 2021, 14(2): 111-128.
[8] Baloch N, Rashid A. Supply chain networks, complexity, and optimization in developing economies: A systematic literature review and meta-analysis: Supply chain networks and complexity: A meta-analysis. South Asian Journal of Operations and Logistics, 2022, 1(1): 14-19.
[9] Lohmer J, Lasch R. Production planning and scheduling in multi-factory production networks: a systematic literature review. International Journal of Production Research, 2021, 59(7): 2028-2054.
[10] Goodarzian F, Shishebori D, Nasseri H, et al. A bi-objective production-distribution problem in a supply chain network under grey flexible conditions. RAIRO-Operations Research, 2021, 55(3): 1971-2000.
[11] Xu W, Song P. Integrated optimisation for production capacity, raw material ordering and production planning under time and quantity uncertainties based on two case studies. Operational Research, 2022, 22(3): 2343-2371.
[12] Saputro T E, Figueira G, Almada-Lobo B. Integrating supplier selection with inventory management under supply disruptions. International Journal of Production Research, 2021, 59(11): 3304-3322.
[13] Zhou B, He Z. A novel hybrid-load AGV for JIT-based sustainable material handling scheduling with time window in mixed-model assembly line. International Journal of Production Research, 2023, 61(3): 796-817.
[14] Jung H. An optimal charging and discharging scheduling algorithm of energy storage system to save electricity pricing using reinforcement learning in urban railway system. Journal of Electrical Engineering & Technology, 2022, 17(1): 727-735.

Downloads: 14589
Visits: 320406

All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.