1.福州大学电气工程与自动化学院,福建省福州市 350108
2.武汉大学电气与自动化学院,湖北省武汉市 430072
张逸(1984—),男,博士,教授,主要研究方向:电网与工业用户电能质量分析、电力数据应用。E-mail:zhangyi@fzu.edu.cn
孙守铨(1999—),男,硕士研究生,主要研究方向:钢铁企业功率预测。E-mail:ssq_zb@163.com
廖思阳(1989—),男,通信作者,博士,教授,博士生导师,主要研究方向:电力系统运行与控制、工业需求响应与新能源并网技术。E-mail:liaosiyang@whu.edu.cn
收稿:2025-05-26,
修回:2025-12-04,
网络首发:2026-03-16,
纸质出版:2026-11-25
移动端阅览
张逸,孙守铨,廖思阳等.基于物理引导图神经网络的短流程钢铁企业负荷预测[J/OL].电力系统自动化, http://dx.doi.org/10.7500/AEPS20250526001.
ZHANG Yi,SUN Shouquan,LIAO Siyang,et al.Load Forecasting of Short-process Iron and Steel Enterprise Based on Physically-guided Graph Neural Network[J/OL].Automation of Electric Power Systems, http://dx.doi.org/10.7500/AEPS20250526001.
张逸,孙守铨,廖思阳等.基于物理引导图神经网络的短流程钢铁企业负荷预测[J/OL].电力系统自动化, http://dx.doi.org/10.7500/AEPS20250526001. DOI:
ZHANG Yi,SUN Shouquan,LIAO Siyang,et al.Load Forecasting of Short-process Iron and Steel Enterprise Based on Physically-guided Graph Neural Network[J/OL].Automation of Electric Power Systems, http://dx.doi.org/10.7500/AEPS20250526001. DOI:
钢铁企业负荷的不确定性给负荷预测精度带来挑战,且现有仅采用数据驱动方法的研究忽略了生产过程中物质流与能量流之间的传递耦合规律。为此,文中提出一种基于物理引导图神经网络的钢铁企业负荷预测方法。首先,基于传输过程衰减机理构建物理信息基函数,以准确反映前后相连工序间的匹配与衔接关系;其次,将钢铁企业网络结构转化为预测模型所适用的图数据;然后,设计物理引导图神经网络的更新策略,在神经网络边更新中将物理信息基函数作为权重更新函数,构建具有一定物理可解释性的负荷预测模型;最后,以国内某钢铁企业为例进行算例分析。结果表明,所提模型方法通过融合物理机理可有效应对不确定性,提高预测准确度。
The uncertainty of load in the iron and steel industry poses challenges to the accuracy of load forecasting. Moreover
existing studies that merely adopt data-driven methods ignore the transfer and coupling laws between material flow and energy flow in the production process. To address this issue
this paper proposes a load forecasting method of the iron and steel industry based on physically-guided graph neural networks. First
physics-informed basis functions are constructed based on the attenuation mechanism of transmission processes to accurately characterize the matching and coupling relationships between sequentially connected working procedures. Second
the network structure of iron and steel enterprises is converted into graph data applicable to the forecasting model. Third
an update strategy for the physically-guided graph neural network is designed. In the edge update of the neural network
the physics-informed basis function is adopted as the weight update function to build a load forecasting model with certain physical interpretability. Finally
a case study is carried out taking a domestic iron and steel enterprise as a case. The results show that the proposed model can effectively handle uncertainty and improve forecasting accuracy by integrating physical mechanisms.
Ruan J Q , Xu Z , Su H . Towards interdisciplinary integration of electrical engineering and earth science [J]. Nature Reviews Electrical Engineering , 2024 , 1 ( 5 ): 278 - 279 .
华昊辰 , 张洲赫 , 邹奕群 , 等 . 新型电力系统需求侧灵活性资源低碳协同优化研究综述 [J]. 电力建设 , 2025 , 46 ( 6 ): 60 - 75 .
Hua Haochen , Zhang Zhouhe , Zou Yiqun , et al . Review of low-carbon co-optimization research on demand-side flexibility resources for new power systems [J]. Electric Power Construction , 2025 , 46 ( 6 ): 60 - 75 .
罗仁杰 , 鞠家鑫 , 李知艺 . 基于混合数学形态学的电力系统灵活性需求分解与运行优化 [J]. 电力系统自动化 , 2025 , 49 ( 11 ): 37 - 47 .
Luo Renjie , Ju Jiaxin , Li Zhiyi . Flexibility demand decomposition and operation optimization for power system based on hybrid mathematical morphology [J]. Automation of Electric Power Systems , 2025 , 49 ( 11 ): 37 - 47 .
Bracale A , Caramia P , De Falco P , et al . A multivariate approach to probabilistic industrial load forecasting [J]. Electric Power Systems Research , 2020 , 187 : 106430 .
Eren Y , KüçüKdemiral İ . A comprehensive review on deep learning approaches for short-term load forecasting [J]. Renewable and Sustainable Energy Reviews , 2024 , 189 : 114031 .
陈振宇 , 刘金波 , 李晨 , 等 . 基于LSTM与XGBoost组合模型的超短期电力负荷预测 [J]. 电网技术 , 2020 , 44 ( 2 ): 614 - 620 .
Chen Zhenyu , Liu Jinbo , Li Chen , et al . Ultra short-term power load forecasting based on combined LSTM-XGBoost model [J]. Power System Technology , 2020 , 44 ( 2 ): 614 - 620 .
吴定会 , 张子恒 , 范俊岩 . 计及多场景因素的钢铁企业多介质能源计划优化研究 [J]. 控制工程 , 2025 , 32 ( 1 ): 12 - 20 .
Wu Dinghui , Zhang Ziheng , Fan Junyan . Research on optimization of multi-medium energy plan of iron and steel enterprises considering multi-scene factors [J]. Control Engineering of China , 2025 , 32 ( 1 ): 12 - 20 .
张谦 , 蒋永梅 , 虞攀峰 . 基于工业集群负荷DR潜力评估的电力市场策略研究 [J]. 农村电气化 , 2025 ( 2 ): 61 - 66 .
Zhang Qian , Jiang Yongmei , Yu Panfeng . Research on power market strategy based on industrial cluster load DR potential assessment [J]. Rural Electrification , 2025 ( 2 ): 61 - 66 .
张效华 . 钢铁企业电力负荷预测方法综述 [C]// 第十四届中国钢铁年会 , 2023 .
Zhang Xiaohua . A review of load forecasting in iron and steel plants [C]// The 14th China Iron and Steel Annual Conference , 2023 .
梁青艳 , 孙彦广 . 钢铁企业电力负荷动态预测建模问题的应用研究 [J]. 科学技术与工程 , 2018 , 18 ( 16 ): 44 - 54 .
Liang Qingyan , Sun Yanguang . Investigation on application of electricity load dynamic forecast modelling in iron and steel enterprises [J]. Science Technology and Engineering , 2018 , 18 ( 16 ): 44 - 54 .
胡宏涛 , 吴定会 , 陆申鑫 , 等 . 基于深度强化学习的钢铁企业能源计划配置 [J]. 控制工程 , 2024 , 31 ( 9 ): 1552 - 1560 .
Hu Hongtao , Wu Dinghui , Lu Shenxin , et al . Energy planning and allocation of iron and steel enterprises based on deep reinforcement learning [J]. Control Engineering of China , 2024 , 31 ( 9 ): 1552 - 1560 .
徐云聪 , 张逸 , 林才华 , 等 . 基于电气特性-物理参数耦合的交流电弧炉模型 [J]. 电工技术学报 , 2024 , 39 ( 6 ): 1643 - 1655 .
Xu Yuncong , Zhang Yi , Lin Caihua , et al . AC arc furnace model based on coupling of electrical and physical parameters [J]. Transactions of China Electrotechnical Society , 2024 , 39 ( 6 ): 1643 - 1655 .
徐吉 , 信自成 , 兰模 , 等 . 冶金机理与贝叶斯优化XGBoost融合的VD炉精炼终点钢液温度预测 [J]. 工程科学与技术 , 2024 , 56 ( 6 ): 63 - 72 .
Xu Ji , Xin Zicheng , Lan Mo , et al . Predicting endpoint temperature of molten steel in VD furnace refining process using metallurgical mechanism and Bayesian optimization XGBoost [J]. Advanced Engineering Sciences , 2024 , 56 ( 6 ): 63 - 72 .
魏子梁 , 郑庆荣 , 汤卓凡 , 等 . 基于仿真模型库的规模化空调负荷需求响应潜力快速评估方法 [J]. 电力需求侧管理 , 2025 , 27 ( 4 ): 71 - 77 .
Wei Ziliang , Zheng Qingrong , Tang Zhuofan , et al . Rapid evaluation method of large-scale air conditioning demand response potential based on simulation model library [J]. Power Demand Side Management , 2025 , 27 ( 4 ): 71 - 77 .
Baur L , Ditschuneit K , Schambach M , et al . Explainability and interpretability in electric load forecasting using machine learning techniques—a review [J]. Energy and AI , 2024 , 16 : 100358 .
Aasim , Singh S N , Mohapatra A . Data driven day-ahead electrical load forecasting through repeated wavelet transform assisted SVM model [J]. Applied Soft Computing , 2021 , 111 : 107730 .
亓晓燕 , 刘恒杰 , 侯秋华 , 等 . 融合LSTM和SVM的钢铁企业电力负荷短期预测 [J]. 山东大学学报(工学版) , 2021 , 51 ( 4 ): 91 - 98 .
Qi Xiaoyan , Liu Hengjie , Hou Qiuhua , et al . Short-term load forecasting of iron and steel industry area based on combination model of SVM and LSTM [J]. Journal of Shandong University (Engineering Science) , 2021 , 51 ( 4 ): 91 - 98 .
Fan G F , Zhang L Z , Yu M , et al . Applications of random forest in multivariable response surface for short-term load forecasting [J]. International Journal of Electrical Power & Energy Systems , 2022 , 139 : 108073 .
李彬 , 张雨蒙 , 周照钒 . 基于LSTM-Transformer的钢铁工业用户调节潜力预测与优化 [J]. 电力系统自动化 , 2026 , 50 ( 5 ): 54 - 62 .
Li Bin , Zhang Yumeng , Zhou Zhaofan . LSTM-transformer based prediction and optimization for user regulation potential of iron and steel industry [J]. Automation of Electric Power Systems , 2026 , 50 ( 5 ): 54 - 62 .
张思 , 李洋 , 王波 , 等 . 基于ICEEMDAN的钢铁行业负荷预测方法 [J]. 浙江电力 , 2024 , 43 ( 5 ): 27 - 34 .
Zhang Si , Li Yang , Wang Bo , et al . A load forecasting method for steel industry based on ICEEMDAN [J]. Zhejiang Electric Power , 2024 , 43 ( 5 ): 27 - 34 .
郝飞 , 解凯 , 孔亮 , 等 . 基于负荷特性的钢铁企业超短期电力负荷预测 [J]. 有色冶金设计与研究 , 2021 , 42 ( 3 ): 28 - 33 .
Hao Fei , Jie Kai , Kong Liang , et al . Ultra short term load forecasting of iron & steel enterprises based on load characteristics [J]. Nonferrous Metals Engineering & Research , 2021 , 42 ( 3 ): 28 - 33 .
Fan C D , Nie S H , Xiao L Y , et al . Short-term industrial load forecasting based on error correction and hybrid ensemble learning [J]. Energy and Buildings , 2024 , 313 : 114261 .
Acikgoz H . A novel approach based on integration of convolutional neural networks and deep feature selection for short-term solar radiation forecasting [J]. Applied Energy , 2022 , 305 : 117912 .
贺东风 , 胡正彪 . 钢铁制造流程物质流与能量流表征与协同评价 [J]. 钢铁 , 2023 , 58 ( 11 ): 32 - 42 .
He Dongfeng , Hu Zhengbiao . Characterization and collaborative evaluation of material flow and energy flow in steel manufacturing process [J]. Iron and Steel , 2023 , 58 ( 11 ): 32 - 42 .
张逸 , 孙守铨 , 廖思阳 , 等 . 基于机理数据混合驱动的钢铁企业生产短流程功率模型 [J]. 中国电机工程学报 , 2025 , 45 ( 18 ): 7098 - 7110 .
Zhang Yi , Sun Shouquan , Liao Siyang , et al . A short process power model for steel production based on mechanism data hybrid drive [J]. Proceedings of the CSEE , 2025 , 45 ( 18 ): 7098 - 7110 .
谷鹏飞 , 孙宏斌 , 薛屹洵 , 等 . 考虑煤气调度的钢铁园区能量流-物质流协同优化方法 [J]. 电力自动化设备 , 2024 , 44 ( 9 ): 32 - 40 .
Gu Pengfei , Sun Hongbin , Xue Yixun , et al . Collaborative optimization method of energy and material flows in iron and steel park considering gas scheduling [J]. Electric Power Automation Equipment , 2024 , 44 ( 9 ): 32 - 40 .
付玲 , 佘玲娟 , 颜镀镭 , 等 . 基于内嵌物理信息与注意力机制BiLSTM神经网络的臂架系统疲劳损伤预测模型 [J]. 机械工程学报 , 2024 , 60 ( 13 ): 205 - 215 .
Fu Ling , She Lingjuan , Yan Dulei , et al . Fatigue damage prediction framework of the boom system based on embedded physical information and attention mechanism BiLSTM neural network [J]. Journal of Mechanical Engineering , 2024 , 60 ( 13 ): 205 - 215 .
Park J Y , Park J K . Physics-induced graph neural network: an application to wind-farm power estimation [J]. Energy , 2019 , 187 : 115883 .
Guo F Z , Li A , Yue B , et al . Improving the out-of-sample generalization ability of data-driven chiller performance models using physics-guided neural network [J]. Applied Energy , 2024 , 354 : 122190 .
Li H R , Gao B , Miao L , et al . Multiphysics structured eddy and thermography defects current diagnostics system in moving mode [J]. IEEE Transactions on Industrial Informatics , 2020 , 17 ( 4 ): 2566 - 2578 .
Alessandro T , Andrea P , Luca L , et al . A Bayesian defect-based physics-guided neural network model for probabilistic fatigue endurance limit evaluation [J]. Computer Methods in Applied Mechanics and Engineering , 2024 , 418 : 116521 .
陆彪 . 钢铁企业能流物流解析及混合建模与集成 [D]. 南京 : 东南大学 , 2017 .
Lu Biao . Hybrid modeling and integration of iron and steel enterprise base on energy flow and material flow analysis [D]. Nanjing : Southeast University , 2017 .
郦秀萍 , 杜学强 , 孙彦广 , 等 . 钢铁制造流程能量流网络的有向网分析及建模 [J]. 中国冶金 , 2024 , 34 ( 12 ): 45 - 53 .
Li Xiuping , Du Xueqiang , Sun Yanguang , et al . Analysis and modeling of energy flow network for steel manufacturing process based on directed network [J]. China Metallurgy , 2024 , 34 ( 12 ): 45 - 53 .
吴博 , 梁循 , 张树森 , 等 . 图神经网络前沿进展与应用 [J]. 计算机学报 , 2022 , 45 ( 1 ): 35 - 68 .
Wu Bo , Liang Xun , Zhang Shusen , et al . Progress and application of graph neural network frontier [J]. Chinese Journal of Computers , 2022 , 45 ( 1 ): 35 - 68 .
刘德澳 , 董明刚 , 叶威 , 等 . 基于图神经网络的非均匀风电场功率预测方法 [J]. 太阳能学报 , 2023 , 44 ( 11 ): 279 - 286 .
Liu Deao , Dong Minggang , Ye Wei , et al . Power prediction method of non-uniform wind farm layout based on graph neural network [J]. Acta Energiae Solaris Sinica , 2023 , 44 ( 11 ): 279 - 286 .
国家能源局 . 关于推进电能替代的指导意见 [EB/OL].[ 2025-02-25 ]. http://wwwnea.gov.cn2016-05/25/c135387453.htm http://wwwnea.gov.cn2016-05/25/c135387453.htm .
National Energy Administration . Guiding opinions on promoting the substitution of electric energy [EB/OL]. [ 2025-02-25 ]. http://wwwnea.gov.cn2016-05/25/c135387453.htm http://wwwnea.gov.cn2016-05/25/c135387453.htm .
李传民 , 王新东 , 习小军 , 等 . 钢铁厂绿色化智能化高效化总体设计技术及应用 [J]. 钢铁 , 2025 , 60 ( 7 ): 41 - 48 .
Li Chuanmin , Wang Xindong , Xi Xiaojun , et al . Overall design technology and application of green intelligent and efficient in steel plants [J]. Iron and Steel , 2025 , 60 ( 7 ): 41 - 48 .
殷瑞钰 . 流程型制造流程的本质与共性规律 [J]. 钢铁 , 2023 , 58 ( 2 ): 1 - 7 .
Yin Ruiyu . Essence and common law of a process oriented manufacturing process [J]. Iron and Steel , 2023 , 58 ( 2 ): 1 - 7 .
崔志峰 , 上官方钦 , 马文略 , 等 . 双碳背景下中国钢铁行业未来发展趋势探讨 [J]. 工程科学学报 , 2025 , 47 ( 4 ): 862 - 874 .
Cui Zhifeng , Shangguan Fangqin , Ma Wenlue , et al . Discussion on the future development trend of China’s iron and steel industry under the background of double carbon [J]. Chinese Journal of Engineering , 2025 , 47 ( 4 ): 862 - 874 .
殷瑞钰 . 冶金流程工程学 [M]. 北京 : 冶金工业出版社 , 2009 .
Yin Ruiyu . Metallurgical process engineering [M]. Beijing : Metallurgical Industry Press , 2009 .
丁婧伊 , 金嘉晖 , 杨丰赫 , 等 . 基于云边协作的工业互联网排产方法:以钢铁热轧生产为例 [J]. 电子学报 , 2024 , 52 ( 9 ): 2988 - 2999 .
Ding Jingyi , Jin Jiahui , Yang Fenghe , et al . Industrial internet scheduling method based on cloud-edge collaboration: a case study of steel hot rolling [J]. Acta Electronica Sinica , 2024 , 52 ( 9 ): 2988 - 2999 .
Xin Z C , Zhang J S , Zhang J G , et al . Mathematical modelling and plant trial on slagging regime in a ladle furnace for high-efficiency desulphurization [J]. Ironmaking Steelmaking , 2021 , 48 ( 9 ): 1123 - 1132 .
徐宪东 , 单文亮 , 蔡云帆 , 等 . 流程工业综合能源系统用电灵活性综述 [J]. 天津大学学报(自然科学与工程技术版) , 2021 , 54 ( 12 ): 1212 - 1220 .
Xu Xiandong , Shan Wenliang , Cai Yunfan , et al . A perspective on electricity flexibility of integrated energy systems in processing industries [J]. Journal of Tianjin University (Science and Technology) , 2021 , 54 ( 12 ): 1212 - 1220 .
0
浏览量
0
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621