特高压直流输电系统的损耗计量是国家电网公司实现节能降耗的重要依据，其数字孪生模型还有助于实现“碳达峰、碳中和”的发展目标。而现有特高压直流系统的数字化水平难以支撑降损等研究的开展，因此本文提出了面向物理空间全映射的数字孪生模型构建算法，主要内容如下。
首先，研究了特高压直流输电系统及其损耗计量。以特高压直流输电系统的拓扑结构和损耗等效电路为基础，研究了损耗计量的组成及子损耗特性。最后结合特高压直流输电系统损耗计量数字空间先验数据集，分析了数据间的强耦合作用，为后续提出基于深度学习模型构建数字孪生模型奠定了基础。
其次，提出了基于深度学习模型构建的损耗计量数字孪生模型。针对数字空间先验数据的强耦合性和长序列性，优选出 LSTM 模型，进一步地提出了改进Res-LSTM 损耗计量数字孪生模型。实验结果表明，在总损耗计算模式和子损耗序列计算模式下，Res-LSTM 模型相比于 LSTM 模型分别降低了 37.57%和 11.8%的 MSE 误差，均优于其他深度学习模型。
随后，提出了面向物理空间全映射数字孪生模型建模与应用算法。针对常规深度学习模型及建模方式的缺陷，提出了全新面向物理空间的全映射数字孪生DTformer 模型建模与应用架构；在此基础上，提出了 DTformer 模型的详细网络结构，以及该模型的全映射建模和应用软策略，提高了数字孪生模型的精确度、应对数据缺失的能力和反演能力。实验结果表明，DTformer 模型的所有误差指标都远远优于常规的深度学习算法。以 MSE 误差为例，最高降低了 92.55%的误差，最低降低了 63.79%的误差。
最后，提出了数字孪生模型的轻量化改进算法。结合轻量模块和架构知识蒸馏两个理论基础，优选跨通道特征交互轻量模块，并构建架构知识蒸馏小模型，将知识蒸馏至轻量模块构建的小架构模型。实验结果表明，改进轻量化算法能实现最快的运算速度，相比于DTformer模型快了7.97倍，相比于蒸馏后的DTformer mini 模型快了 1.28 倍，损失的精度在可接受范围内。
本文算法实现了面向物理空间全映射的数字孪生模型的建模与损耗计量应用，为后续降损优化等研究和应用奠定了良好的基础。

[1] Chen Zhengxi, Wang Chen, Li Jun, et al. Conductor selection and economic analysis of D.R. congo–guinea ±800 kV UHVDC transmission project[J]. Global Energy Interconnection, 2020,3(4):385-397.DOI:https://doi.org/10.1016/j.gloei.2020.10.007.

[2] 国家电网有限公司. 电网业务领域/特高压[EB/OL]. [3月19日]. http://www.sgcc.com.cn/html/sgcc_main/col2017041259/column_2017041259_1.shtml?childColumnId=2017041259.

[3] Wu Chenxi, Zhang Hongli, Hou Yuqiang, et al. Analysis of Binjin UHVDC restart failure and relevant suggestions on secure and stable operation of power grid[J]. Energy Procedia, 2018,145:452-457.DOI:https://doi.org/10.1016/j.egypro.2018.04.086.

[4] F. M. Kasangala, G. Atkinson-Hope. Electrical energy losses and costs evaluation of HVDC and UHVDC transmission lines: 2013 Proceedings of the 10th Industrial and Commercial Use of Energy Conference, 2013[C].

[5] 杨帆, 吴涛, 廖瑞金, 等. 数字孪生在电力装备领域中的应用与实现方法[J]. 高电压技术, 2021,47(05):1505-1521.DOI:10.13336/j.1003-6520.hve.20210456.

[6] 齐波, 张鹏, 张书琦, 等. 数字孪生技术在输变电设备状态评估中的应用现状与发展展望[J]. 高电压技术, 2021,47(05):1522-1538.DOI:10.13336/j.1003-6520.hve.20210093.

[7] A. Fuller, Z. Fan, C. Day, et al. Digital Twin: Enabling Technologies, Challenges and Open Research[J]. IEEE Access, 2020,8:108952-108971.DOI:10.1109/ACCESS.2020.2998358.

[8] 刘蔚然, 陶飞, 程江峰, 等. 数字孪生卫星:概念、关键技术及应用[J]. 计算机集成制造系统, 2020,26(03):565-588.DOI:10.13196/j.cims.2020.03.001.

[9] H. Laaki, Y. Miche, K. Tammi. Prototyping a Digital Twin for Real Time Remote Control Over Mobile Networks: Application of Remote Surgery[J]. IEEE Access, 2019,7:20325-20336.DOI:10.1109/ACCESS.2019.2897018.

[10] 李亚楠. 智能网联汽车数字孪生测试理论和技术研究[D]. 吉林大学, 2020.

[11] 刘理想. 电弧增材制造堆积形貌研究及堆积过程数字孪生模拟[D]. 南华大学, 2020.

[12] P. Jain, J. Poon, J. P. Singh, et al. A Digital Twin Approach for Fault Diagnosis in Distributed Photovoltaic Systems[J]. IEEE Transactions on Power Electronics, 2020,35(1):940-956.DOI:10.1109/TPEL.2019.2911594.

[13] C. Gehrmann, M. Gunnarsson. A Digital Twin Based Industrial Automation and Control System Security Architecture[J]. IEEE Transactions on Industrial Informatics, 2020,16(1):669-680.DOI:10.1109/TII.2019.2938885.

[14] 陶飞, 刘蔚然, 张萌, 等. 数字孪生五维模型及十大领域应用[J]. 计算机集成制造系统, 2019,25(01):1-18.DOI:10.13196/j.cims.2019.01.001.

[15] 王乐, 赵彪, 韩丰, 等. 直流输电线路损耗研究[J]. 电力建设, 2008,29(9):4.

[16] M. M. Rathore, S. A. Shah, D. Shukla, et al. The Role of AI, Machine Learning, and Big Data in Digital Twinning: A Systematic Literature Review, Challenges, and Opportunities[J]. IEEE Access, 2021,9:32030-32052.DOI:10.1109/ACCESS.2021.3060863.

[17] 梁涵卿, 邬雄, 梁旭明. 特高压交流和高压直流输电系统运行损耗及经济性分析[J]. 高电压技术, 2013,39(03):630-635.

[18] 周国伟, 顾用地, 周建平, 等. 特高压直流输电系统非特征谐波分析[J]. 电力系统保护与控制, 2016,44(03):122-128.

[19] 韩民晓, 文俊, 徐永海. 高压直流输电原理与运行[M]. 高压直流输电原理与运行, 2009.

[20] F. Yu, Y. Wu, J. Guo. Research on the resonance characteristics of ±1100kV UHVDC transmission system: The 16th IET International Conference on AC and DC Power Transmission (ACDC 2020), 2020[C].

[21] Sanders Jan A., Verhulst Ferdinand, Murdock James. Averaging methods in nonlinear dynamical systems[M]. Springer, 2007.

[22] Jatskevich Juri, Pekarek Steven D., Davoudi Ali. Fast procedure for constructing an accurate dynamic average-value model of synchronous machine-rectifier systems[J]. IEEE Transactions on Energy Conversion, 2006,21(2):435-441.

[23] Huang Zhenyu, Yang Bo, Kosterev Dmitry. Benchmarking of planning models using recorded dynamics: 2009 IEEE/PES Power Systems Conference and Exposition, 2009[C].

[24] 潘学萍, 温荣超, 鞠平, 等. 双馈风电机组网侧控制器参数辨识的频域方法[J]. 电网技术, 2015,39(3):5.

[25] 汪应春, 唐登平, 余鹤, 等. 一种跨省输电线路线损智能管理与决策方法及系统[Z]. 2020.

[26] 刘俊, 潘玲玲, 王勇, 等. 一种基于电量数据的输电线路电阻参数估计方法及装置[Z]. 2020.

[27] B. Chen, K. Xiang, L. Yang, et al. Theoretical Line Loss Calculation of Distribution Network Based on the Integrated Electricity and Line Loss Management System: 2018 China International Conference on Electricity Distribution (CICED), 2018[C].

[28] 姚力, 韩霄汉, 胡瑛俊, 等. 一种用于柔性直流输电系统的能效计量方法[Z]. 2016.

[29] Glaessgen Edward, Stargel David. The digital twin paradigm for future NASA and US Air Force vehicles: 53rd AIAA/ASME/ASCE/AHS/ASC structures, structural dynamics and materials conference 20th AIAA/ASME/AHS adaptive structures conference 14th AIAA, 2012[C].

[30] Kraft Edward M. The air force digital thread/digital twin-life cycle integration and use of computational and experimental knowledge: 54th AIAA aerospace sciences meeting, 2016[C].

[31] Negri Elisa, Fumagalli Luca, Macchi Marco. A review of the roles of digital twin in CPS-based production systems[J]. Procedia Manufacturing, 2017,11:939-948.

[32] Boschert Stefan, Rosen Roland. Digital twin—the simulation aspect[M]//Mechatronic futures. Springer, 2016:59-74.

[33] Rosen Roland, Von Wichert Georg, Lo George, et al. About the importance of autonomy and digital twins for the future of manufacturing[J]. Ifac-papersonline, 2015,48(3):567-572.

[34] Grieves Michael. Digital twin: manufacturing excellence through virtual factory replication[J]. White paper, 2014,1:1-7.

[35] Grieves Michael, Vickers John. Digital twin: Mitigating unpredictable, undesirable emergent behavior in complex systems[M]//Transdisciplinary perspectives on complex systems. Springer, 2017:85-113.

[36] Uhlemann Thomas H-J, Lehmann Christian, Steinhilper Rolf. The digital twin: Realizing the cyber-physical production system for industry 4.0[J]. Procedia Cirp, 2017,61:335-340.

[37] Tao Fei, Cheng Jiangfeng, Qi Qinglin, et al. Digital twin-driven product design, manufacturing and service with big data[J]. The International Journal of Advanced Manufacturing Technology, 2018,94(9):3563-3576.

[38] 陶飞, 程颖, 程江峰, 等. 数字孪生车间信息物理融合理论与技术[J]. 计算机集成制造系统, 2017,23(08):1603-1611.DOI:10.13196/j.cims.2017.08.001.

[39] 陶飞, 张萌, 程江峰, 等. 数字孪生车间——一种未来车间运行新模式[J]. 计算机集成制造系统, 2017,23(01):1-9.DOI:10.13196/j.cims.2017.01.001.

[40] Tao Fei, Sui Fangyuan, Liu Ang, et al. Digital twin-driven product design framework[J]. International Journal of Production Research, 2019,57(12):3935-3953.

[41] Tao Fei, Zhang Meng, Liu Yushan, et al. Digital twin driven prognostics and health management for complex equipment[J]. Cirp Annals, 2018,67(1):169-172.

[42] Qi Qinglin, Tao Fei. Digital twin and big data towards smart manufacturing and industry 4.0: 360 degree comparison[J]. Ieee Access, 2018,6:3585-3593.

[43] 庄存波, 刘检华, 熊辉, 等. 产品数字孪生体的内涵,体系结构及其发展趋势[J]. 计算机集成制造系统, 2017,23(4):16.

[44] 于勇, 胡德雨, 戴晟, 等. 数字孪生在工艺设计中的应用探讨[J]. 航空制造技术, 2018,61(18):8.

[45] 王岭. 基于数字孪生的航空发动机低压涡轮单元体对接技术研究[J]. 计算机测量与控制, 2018,26(10):286-290.DOI:10.16526/j.cnki.11-4762/tp.2018.10.061.

[46] 张新生. 基于数字孪生的车间管控系统的设计与实现[D]. 郑州大学, 2018.

[47] 戴晟, 赵罡, 于勇, 等. 数字化产品定义发展趋势:从样机到孪生[J]. 计算机辅助设计与图形学学报, 2018,30(08):1554-1562.

[48] 刘青, 刘滨, 王冠, 等. 数字孪生的模型、问题与进展研究[J]. 河北科技大学学报, 2019,40(01):68-78.

[49] 周毅, 周良才, 沈颖平. 基于数字孪生的华东电网安控系统虚拟建模及实现[J]. 电器与能效管理技术, 2021(06):47-51.DOI:10.16628/j.cnki.2095-8188.2021.06.008.

[50] 卢晓光, 李凤格. 基于数字孪生的风机实时载荷预估研究[J]. 机械与电子, 2021,39(06):24-28.

[51] Ren S., He K., Girshick R., et al. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2017,39(6):1137-1149.

[52] Lin Tsung-Yi, Goyal Priya, Girshick Ross, et al. Focal loss for dense object detection: Proceedings of the IEEE international conference on computer vision, 2017[C].

[53] Tian Zhi, Shen Chunhua, Chen Hao, et al. Fcos: Fully convolutional one-stage object detection: Proceedings of the IEEE/CVF international conference on computer vision, 2019[C].

[54] W. Danilczyk, Y. L. Sun, H. He. Smart Grid Anomaly Detection using a Deep Learning Digital Twin: 2020 52nd North American Power Symposium (NAPS), 2021[C].

[55] M. Zhou, Z. Hou, J. Liu, et al. Digital Twin-based Automatic Train Regulation for Integration of Dispatching and Control: 2021 IEEE 1st International Conference on Digital Twins and Parallel Intelligence (DTPI), 2021[C].

[56] Krizhevsky Alex, Sutskever Ilya, Hinton Geoffrey E. ImageNet classification with deep convolutional neural networks[J]. Commun. ACM, 2017,60(6):84-90.DOI:10.1145/3065386.

[57] J. Wu, Y. Dang, H. Jia, et al. Prediction of Energy Consumption in Digital Twins of Intelligent Factory by Artificial Intelligence: 2021 International Conference on Technology and Policy in Energy and Electric Power (ICT-PEP), 2021[C].

[58] Hochreiter, Sepp, Schmidhuber, et al. Long short-term memory.[J]. Neural Computation, 1997.

[59] Zaremba Wojciech, Sutskever Ilya, Vinyals Oriol. Recurrent neural network regularization[J]. arXiv preprint arXiv:1409.2329, 2014.

[60] Bahdanau Dzmitry, Cho Kyunghyun, Bengio Yoshua. Neural machine translation by jointly learning to align and translate[J]. arXiv preprint arXiv:1409.0473, 2014.

[61] Lai Guokun, Chang Wei-Cheng, Yang Yiming, et al. Modeling long-and short-term temporal patterns with deep neural networks: The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval, 2018[C].

[62] H. V. Dang, H. Tran-Ngoc, T. V. Nguyen, et al. Data-Driven Structural Health Monitoring Using Feature Fusion and Hybrid Deep Learning[J]. IEEE Transactions on Automation Science and Engineering, 2021,18(4):2087-2103.DOI:10.1109/TASE.2020.3034401.

[63] Vaswani Ashish, Shazeer Noam, Parmar Niki, et al. Attention is all you need[C]//: Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, California, USA, 2017. Curran Associates Inc..

[64] Song Huan, Rajan Deepta, Thiagarajan Jayaraman J., et al. Attend and diagnose: Clinical time series analysis using attention models: Thirty-second AAAI conference on artificial intelligence, 2018[C].

[65] Li Shiyang, Jin Xiaoyong, Xuan Yao, et al. Enhancing the locality and breaking the memory bottleneck of transformer on time series forecasting[J]. Advances in Neural Information Processing Systems, 2019,32.

[66] Child Rewon, Gray Scott, Radford Alec, et al. Generating long sequences with sparse transformers[J]. arXiv preprint arXiv:1904.10509, 2019.

[67] Zhou Haoyi, Zhang Shanghang, Peng Jieqi, et al. Informer: Beyond efficient transformer for long sequence time-series forecasting: Proceedings of AAAI, 2021[C].

[68] X. Xie, A. K. Parlikad, R. S. Puri. A Neural Ordinary Differential Equations Based Approach for Demand Forecasting within Power Grid Digital Twins: 2019 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm), 2019[C].

[69] Choudhary Tejalal, Mishra Vipul, Goswami Anurag, et al. A comprehensive survey on model compression and acceleration[J]. Artificial Intelligence Review, 2020,53(7):5113-5155.DOI:10.1007/s10462-020-09816-7.

[70] Iandola Forrest N., Han Song, Moskewicz Matthew W., et al. SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 0.5 MB model size[J]. arXiv preprint arXiv:1602.07360, 2016.

[71] Hu Jie, Shen Li, Sun Gang. Squeeze-and-excitation networks: Proceedings of the IEEE conference on computer vision and pattern recognition, 2018[C].

[72] Fu Jun, Liu Jing, Tian Haijie, et al. Dual attention network for scene segmentation: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019[C].

[73] Hinton Geoffrey, Vinyals Oriol, Dean Jeff. Distilling the knowledge in a neural network[J]. arXiv preprint arXiv:1503.02531, 2015,2(7).

[74] 葛仕明, 赵胜伟, 刘文瑜, 等. 基于深度特征蒸馏的人脸识别[J]. 北京交通大学学报, 2017,41(06):27-33.

[75] 赵胜伟, 葛仕明, 叶奇挺, 等. 基于增强监督知识蒸馏的交通标识分类[J]. 中国科技论文, 2017,12(20):2355-2360.

[76] 邓棋, 雷印杰, 田锋. 用于肺炎图像分类的优化卷积神经网络方法[J]. 计算机应用, 2020,40(01):71-76.

[77] 李坚强, 王成, 黄志超, 等. 基于集成知识蒸馏的肺病分析[J]. 控制理论与应用, 2021,38(01):130-136.

[78] 苏仁恒. 探析特高压直流电对山东电网交流系统的影响[D]. 山东大学, 2018.

[79] 中国电力科学研究院, 西安高压电器研究院有限责任公司, 西安西电电力系统有限公司, 等. 高压直流换流站损耗的确定[S]. 中华人民共和国国家质量监督检验检疫总局;中国国家标准化管理委员会.

[80] 西安西电电力系统有限公司, 西安高压电器研究院有限责任公司, 国网北京经济技术研究院, 等. 高压直流系统用电压源换流器阀损耗 第1部分：一般要求[S]. 中华人民共和国国家质量监督检验检疫总局;中国国家标准化管理委员会.

[81] 娄彦涛, 袁妮, 刘琦. 高压直流输电晶闸管换流阀损耗仿真计算方法研究[J]. 高压电器, 2021,57(01):116-122.DOI:10.13296/j.1001-1609.hva.2021.01.017.

[82] TC IEC. 换流变压器.第1部分:工业用变压器[S]. 国际电工委员会.

[83] 西安高压电器研究所. 高压直流系统交流滤波器[S]. 中华人民共和国国家质量监督检验检疫总局;中国国家标准化管理委员会.

[84] 国家电网公司直流建设分公司. ±800kV及以下换流站干式平波电抗器施工及验收规范[S]. 中华人民共和国住房和城乡建设部;中华人民共和国国家质量监督检验检疫总局.

[85] 西安高压电器研究院有限责任公司, 南方电网技术研究中心. 高压直流输电系统直流滤波器[S]. 中华人民共和国国家质量监督检验检疫总局;中国国家标准化管理委员会.

[86] TC IEC. 电缆.计算额定电流.第1-1部分:额定电流方程(100%负载因数)和损耗计算.总则[S]. 国际电工委员会.

[87] 王昊泽. 跨省输电线路理论线损模型及线损率异常成因研究[D]. 华中科技大学, 2018.

[88] He K., Zhang X., Ren S., et al. Deep Residual Learning for Image Recognition[Z]. Los Alamitos, CA, USA: IEEE Computer Society, 2016: {}, 770-778.

[89] Tay Yi, Dehghani Mostafa, Bahri Dara, et al. Efficient transformers: A survey[J]. arXiv preprint arXiv:2009.06732, 2020.

[90] Liu Ziquan, Cui Yufei, Chan Antoni B. Improve generalization and robustness of neural networks via weight scale shifting invariant regularizations[J]. arXiv preprint arXiv:2008.02965, 2020.

[91] Wang Qilong, Wu Banggu, Zhu Pengfei, et al. ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks[J]. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020:11531-11539.