.png){kind=icon}
返回
Multi-objective operation optimization of a steelmaking process with data analytics modelling and a direct multi-search algorithm
.png){kind=icon}
PRE
AI
DOI：10.1080/0305215X.2025.2573004.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Operation optimization is crucial for enhancing molten steel quality and controlling temperature in steelmaking. However, the extreme thermal conditions, multi-objective coupling and imprecise process mechanisms establish this process as a typical black-box system, severely constraining the effectiveness of conventional optimization methods. Thus, a multi-input multi-output data analytics model is developed using a backpropagation neural network, upon which a multi-objective operation optimization model is constructed. Targeting the model's black-box nature, an improved multi-objective derivative-free optimization algorithm is proposed based on the direct multi-search framework. By integrating a comparison function and a search step design, this approach fully leverages the information from evaluated solutions and iterative process data, thereby improving Pareto front quality. Numerical experiments with industrial data demonstrate the algorithm's competitiveness and effectiveness.
Keyword:
Steelmaking
backpropagation neural network
direct multi-search
multi-objective derivative-free optimization
comparison function
期刊
IF:
2.2
论文数:
109
被引数:
3.8K
机构
引用论文
.png){kind=icon}
分享收藏Prediction model of end-point phosphorus content in BOF steelmaking process based on PCA and BP neural network基于PCA和BP神经网络的BOF炼钢终点磷含量预测模型
PREAI
分享收藏Evolutionary Multi-Objective Bayesian Optimization Based on Multisource Online Transfer Learning
PREAI
分享收藏分享收藏
Derivative-Free Optimization: Lifting Single-Objective to Multi-Objective AlgorithmDejemeppe, C.; Schaus, P.; Deville, Y. 无导数优化：将单目标提升至多目标算法。载于《约束规划中人工智能与运筹学技术的集成》会议论文集，巴塞罗那，西班牙，2015年5月18-22日；主编：Michel, L.；出版商：Springer, 瑞士楚格，2015年；页码：124-140。[Google Scholar] [CrossRef]
PREAI
分享收藏
查看更多内容