.png){kind=icon}
Return
Multi-objective load distribution optimization for hot continuous rolling mills based on an improved differential evolution algorithm
.png){kind=icon}
PRE
AI
DOI：10.1088/2631-8695/ae6975.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In order to improve the enforceability and overall performance of load distribution under strong constraint operation conditions, a multi-objective optimization model of load distribution for multi-stand hot strip mill is proposed. The model combines actual production constraints and is solved by an improved multi-objective algorithm based on differential evolution, which simultaneously optimizes product quality, energy consumption and equipment operation stability. Traditional empirical allocation strategies and existing multi-objective algorithms often struggle with coupling constraints, operational interference and high-dimensional decision variables. These challenges may lead to premature convergence, low proportion of feasible solutions, uneven Pareto frontier distribution and unstable engineering performance. In order to solve these problems, this study integrates thickness control, unit energy consumption, peak power suppression and load stability into a unified optimization framework. It further introduces adaptive parameter adjustment, hybrid mutation strategy and feasibility priority selection mechanism to improve the search efficiency and the feasibility of Pareto solution. The study uses a common subset of multi-station rolling data sets to evaluate the proposed method. The algorithm achieves the Hypervolume (HV) values of 0.824, 0.817 and 0.808 on the stream 1, 3 and 6 of tandem cold mill (TCM), respectively, which is superior to the comparison algorithm. Convergence analysis shows that under typical working conditions, the 200th generation HV value of this method reaches 0.817, with stable convergence and strong optimization ability. From the engineering point of view, the average thickness deviation (MTD) of the three data sets is 0.21 mm, 0.22 mm and 0.24 mm. The corresponding unit energy consumption values are 46.18 kWh ton-1, 46.93 kWh ton-1 and 47.62 kWh ton-1, and the peak power reduction rate is 11.72%-12.48%. The proportion of feasible solutions reaches 93.28%, 91.64% and 89.57% respectively. These results show that the proposed method achieves a balance between quality control, energy efficiency and peak load stability, while maintaining high constraint feasibility and robustness. Overall, the proposed optimization framework provides an effective and practical solution for multi-stand load distribution in hot strip rolling mills. The findings also offer useful insights for multi-objective optimization in rolling processes and intelligent manufacturing systems.
Keywords:
hot continuous rolling mill
multi-objective load distribution
improved differential evolution
Pareto optimality
Journal
E
IF:
1.6
Papers:
2.1K
Citations:
0
Organization
No organization information available
Cited Papers
No cited papers available