Design and Application of a Large-Scale Multi-Objective Dynamic Steel Allocation Optimization Model Integrating Thought Chains
In view of the problems of traditional steel distribution relying on manual experience, poor dynamic adaptability and multi-objective optimization imbalance, a multi-objective dynamic steel distribution optimization model integrating the thinking chain is proposed.The model is based on the scrap steel image data set, expands the data through StyleGAN3 and Improved-Diffusion models, improves robustness by Swin-IR ultra-resolution processing, and adopts Swin-Transformer and SE Attention mechanism optimization YOLOv8 to build a scrap steel grading model;Fusion reinforcement learning and cross-modal feature adaptive fusion technology, establish a multi-objective decision-making system including cost, quality and energy consumption, and combine metallurgical process knowledge map and reaction dynamics model to achieve constraint optimization.Through the dialogue and interactive interface, the optimal steel scheme can be output by entering the target steel type. Experimental verification shows that the control accuracy of the steel composition of the model has been increased by more than 15%, the energy consumption has been reduced by 8%, and the production cost has been reduced by 10%, providing technical support for the intelligent and low-carbon transformation of the steel industry.
References
- Guo Qingcheng. Structural design and optimization of copper-based composite materials based on BP neural network and multi-objective optimization [D]. Xi'an University of Technology, 2024.
- Zhao Yuntao, Deng Xinhui. Attention feature fusion attitude estimation network based on multimodal data [J]. Liquid crystal and display, 2025, 40 (04): 598-606.
- Hao Shengtao. Research on the classification and measurement system and data collection method of steelmaking scrap steel [J]. Shanxi Metallurgy, 2023, 46 (07): 192-194.
- Digital modeling of scrap steel in Zhanjiang Base of Baosteel [J]. Angang Technology, 2022, (06): 11.
- Hayashi S, Tamaki W, TOMOTA Y. Relationship between scrap mixing and steel products for electric furnace [J]. Tetsu-to-Hagane, 2005, 91 (1): 147-149.
- Huang X, Liu X, Liu Y, et al. Element Detection in Scrap Steel Using Portable LIBS and Sparrow Search Algorithm-Kernel Extreme Learning Machine (SSA-KELM)[J]. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2024, 44 (9): 2412-2419.
- Alvarez E A, Trashorras A J G, Cuesta J M S, et al. Steel mill slags energy potential: the case of the steel factory of Arcelor-Mittal in Asturias (Spain)[J]. Clean Technologies and Environmental Policy, 2012, 14: 869-877.