Document Type : Original Article
Authors
1
Department of Industrial Engineering, Faculty of Technology and Engineering, University of Qom, Qom, Iran.
2
Department of Management, Faculty of Management, University of Tehran, Tehran, Iran. Email: amohaghar@ut.ac.ir
3
Department of Industrial Engineering, Faculty of Technology and Engineering, University of Qom, Qom, Iran. Email: mokhtariy@yahoo.com
4
Department of Management, Faculty of Management, University of Qom, Qom, Iran. Email: bozorgmehr.maleki1363@gmail.com
Abstract
Determining production lot sizes and the raw material cutting problem are key issues in many manufacturing industries. These two problems play a significant role in raw material management, production planning, and cost control in industries such as metals, paper, furniture, and aluminum. Determining the lot size relates to decisions about the quantity of each product to produce in different periods, while the raw material cutting problem deals with determining cutting patterns with minimal waste and cost. Given the interdependence of these two problems, integrated decision-making concerning them can significantly impact the reduction of overall system costs and the improvement of operational productivity. This research presents an integrated mathematical model that combines these two problems. In addition to the multi-level and multi-period production structure, the model also considers costs imposed on the system, including order delivery delays, overtime, and raw material procurement. The main innovation of this research lies in its problem-solving approach; instead of using exact methods or common metaheuristic algorithms, a method based on deep reinforcement learning is used to enable decision-making within the large and complex problem space. In this approach, the problem is designed so that the learning method can make the best decisions regarding production quantity, inventory levels, and the selection of cutting patterns according to different system states. This method can adapt to various operational conditions and functions in an integrated manner without needing to decompose the problem into separate parts
Keywords
Main Subjects
)
Send comment about this article