收藏切换
Multi-objective optimization scheduling of steel-industry virtual power plants based on physics-informed neural networks
收藏切换
Feixiang Gong1, Xun Wang2, 3, Ziqian Zhang1, Kaicheng Liu1, Ming Cheng2, 3
Journal of Materials and Metallurgy | 2026, 25(5) : 434 - 442
Less
收藏切换
Journal of Materials and Metallurgy | 2026, 25(5): 434-442
• Special Column: Technological Pathways for Synergistic Decarbonization of the Steel and Power Industries •
Multi-objective optimization scheduling of steel-industry virtual power plants based on physics-informed neural networks
Full
Feixiang Gong1, Xun Wang2, 3, Ziqian Zhang1, Kaicheng Liu1, Ming Cheng2, 3
Affiliations
  • 1.China Electric Power Research Institute Co., Ltd., Beijing 100192, China
  • 2.State Grid Jibei Electric Power Co., Ltd., Beijing 100032, China
  • 3.State Key Laboratory of Intelligent Power Equipment and System, Tianjin 300401, China
Published: 2026-09-10 doi: 10.14186/j.cnki.1671-6620.2026.147
Outline
收藏切换

The large load scale and strong regulation potential of the steel industry can provide flexibility support for industrial virtual power plants under high-proportion renewable energy integration. To address the difficulty of conventional scheduling methods in simultaneously balancing economy, low carbon emissions, and renewable energy accommodation under process constraints in steel production, including minimum production boundaries, ramping limits, daily electricity consumption tasks, and waste-heat utilization coupling, this paper first constructs a typical-day virtual power plant scenario consisting of eight steel enterprises. A multi-objective optimal scheduling model is then established with operating cost, carbon emissions, and renewable energy accommodation as the optimization objectives, while production boundaries, ramping limits, daily electricity consumption tasks, waste-heat power generation coupling, and energy storage state constraints are incorporated into a unified framework. Second, a physics-informed neural network-based optimization method is proposed, in which state reconstruction and a composite loss function are used to achieve joint training for objective optimization and constraint satisfaction. Finally, a typical-day case study with a 24-hour horizon and 1-hour resolution shows that, compared with MOEA/D and RVEA, the proposed method reduces the comprehensive objective value by approximately 3.4% and 4.8%, respectively, and increases the renewable energy accommodation rate by 10.9 and 7.8 percentage points, respectively. The proposed method can provide a reference for low-carbon collaborative scheduling of energy-intensive industrial virtual power plants.

virtual power plant  /  steel enterprise  /  multi-objective optimal dispatch  /  renewable energy consumption  /  physics-informed neural network
宫飞翔, 王珣, 张梓乾, 刘铠诚, 程明. 基于物理信息神经网络的钢铁行业虚拟电厂多目标优化调度. 材料与冶金学报, 2026 , 25 (5) : 434 -442 . DOI: 10.14186/j.cnki.1671-6620.2026.147
Feixiang Gong, Xun Wang, Ziqian Zhang, Kaicheng Liu, Ming Cheng. Multi-objective optimization scheduling of steel-industry virtual power plants based on physics-informed neural networks[J]. Journal of Materials and Metallurgy, 2026 , 25 (5) : 434 -442 . DOI: 10.14186/j.cnki.1671-6620.2026.147
Year 2026 volume 25 Issue 5
57
14
Cite this Article
BibTeX
Article Info
doi: 10.14186/j.cnki.1671-6620.2026.147
  • Receive Date:2026-05-29
  • Online Date:2026-11-13
  • Published:2026-09-10
Article Data
Affiliations
History
  • Received:2026-05-29
Funding
Affiliations
    1.China Electric Power Research Institute Co., Ltd., Beijing 100192, China
    2.State Grid Jibei Electric Power Co., Ltd., Beijing 100032, China
    3.State Key Laboratory of Intelligent Power Equipment and System, Tianjin 300401, China
References
Share
https://castjournals.cast.org.cn/joweb/clyyjxb/EN/10.14186/j.cnki.1671-6620.2026.147
Share to
QR

Scan QR to access full text

Cite this article
BibTeX
Citations
表12种不同金属材料的力学参数
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏
  • BibTeX
  • EndNote
  • RefWorks
  • TxT