Optimal production scheduling and roll change by joint multi-attribute constrained decision-making in a hot rolling process

• Published in: International journal of advanced manufacturing technology Vol. 139; no. 9-10; pp. 4939 - 4957
• Main Authors: Hu, Shangju, Zhang, Xiaohong, Zeng, Jianchao, Gan, Jie, Shi, Guannan, Qiu, Huadong
• Format: Journal Article
• Language: English
• Published: London Springer London 01.08.2025; Springer Nature B.V
• Subjects: Algorithms; CAE) and Design; Computer-Aided Engineering (CAD; Constraints; Costs; Decision making; Efficiency; Engineering; Genetic algorithms; Hot rolling; Industrial and Production Engineering; Literature reviews; Mechanical Engineering; Media Management; Neural networks; Optimization; Original Article; Parameter estimation; Prediction models; Product quality; Production scheduling; Quality control; Reliability; Scheduling; Tabu search; Production scheduling; Multi-attribute constraints; Hot rolling; Roll wear; Roll change
• ISSN: 0268-3768; 1433-3015
• DOI: 10.1007/s00170-025-16082-w

In the hot rolling process, orders have varying quality grades and delivery requirements. Roll changes are essential to maintaining product quality, as wear of the rolls determines when changes are necessary. Frequent roll changes interrupt production and delay delivery. This paper addresses multiple constraints, including quality and delivery, and examines joint decision-making for optimal production scheduling and roll change to balance product quality and delivery time and improve efficiency. A random wear degradation and reliability prediction model is developed, considering differences in slab material and production processes. Second, a decision-making model of production scheduling and roll changes under dual quality and delivery time constraints is constructed, determining the optimal slab sequence and roll change threshold to maximize profit. A hybrid optimization algorithm, combining a genetic algorithm and tabu search, solves the model, improving search efficiency. Numerical experiments using data from 1009 slabs in an actual steel plant validate model’s effectiveness. Results show that the proposed strategy can increase profit by 2.62% and reduce the number of roll changes, ensuring reliability above 0.97 while meeting quality requirements.