温文涛,杨振华,漆乡萌,邓慧

• 1:暨南大学能源电力研究中心

摘要(Abstract)：

在灵活调峰的背景下，为适应直接空冷机组负荷动态变化与环境因素干扰，提出一种在线学习的神经网络方法对直接空冷机组背压进行预测。首先，对历史数据进行清洗，通过Spearman相关性分析确定影响运行背压的低冗余重要特征。接着，采用Hammerstein模型对背压进行模型参数在线辨识。同时，采用长短记忆神经网络和注意力机制建立直接空冷机组背压预测模型，使用在线学习的方式对模型进行更新。实验表明：该模型在预测未来1h内不同时间跨度的背压绝对百分比误差（MAPE）低于9%，并在预测30s内的背压MAPE低于1%。最后，在实际电厂系统中验证模型能够在实际应用中稳定运行。本研究的成果为直接空冷机组背压实时预测提供了有效的方法，这对于灵活调峰直接空冷机组的运行和管理具有重要的意义。

关键词(KeyWords)： 直接空冷机组;背压预测;在线学习;注意力机制;长短期记忆神经网络

Abstract:

Keywords:

基金项目(Foundation): 暨南大学特色新工科起点建设项目（G20200019251）;; 白城发电公司项目（410011JX202000244） ~~

作者(Author): 温文涛,杨振华,漆乡萌,邓慧

DOI: 10.19666/j.rlfd.202309150

参考文献(References)：

• [1]中国电力企业联合会.中电联发布2023年上半年全国电力供需形势分析预测报告[R/OL].(2023-07-26)[2023-07-29]. http://www.chinapower.com. cn/xw/zyxw/20230726/210472.html.China Electricity Union.China Electricity Union released the first half of 2023 national electricity supply and demand situation analysis forecast report[R/OL].(2023-07-26)[2023-07-29]. http://www.chinapower.com. cn/xw/zyxw/20230726/210472.html.
• [2]国家发展改革委.国家能源局关于发展全国煤电机组改造升级的通知[R/OL].(2021-10-29)[2023-07-29].https://www.gov.cn/zhengce/zhengceku/2021-11/03content_5648562.htm.National Development and Reform Commission. NEA on the development of national coal power unit transformation and upgrading notice[R/OL].(2021-10-29)[2023-07-29]. https://www.gov.cn/zhengce/zhengceku/2021-11/03/content_5648562.htm.
• [3] PAN S Y, SNYDER S W, PACKMAN A I, et al. Cooling water use in thermoelectric power generation and its associated challenges for addressing water-energy nexus[J]. Water-Energy Nexus, 2018, 1(1):26-41.
• [4]曾世权,开平安,庄志宝,等.灵活运行空冷机组频域自适应背压控制与优化[J].热力发电, 2023, 52(5):115-121.ZENG Shiquan, KAI Ping’an, ZHUANG Zhibao, et al.Adaptive back pressure control and optimization in frequency domain for flexible operation air cooling unit[J]. Thermal Power Generation, 2023, 52(5):115-121.
• [5] LI J, BAI Y, LI B. Operation of air cooled condensers for optimised backpressure at ambient wind[J]. Applied Thermal Engineering, 2018, 128:1340-1350.
• [6]李健.直接空冷机组建模与优化控制研究[D].北京:华北电力大学, 2019:1.LI Jian. Research on modeling and optimal control of direct air cooling unit[D]. Beijing:North China Electric Power University, 2019:1.
• [7] LI X, WANG N, WANG L, et al. A data-driven model for the air-cooling condenser of thermal power plants based on data reconciliation and support vector regression[J].Applied Thermal Engineering, 2018, 129:1496-1507.
• [8] HUNDI P, SHAHSAVARI R. Comparative studies among machine learning models for performance estimation and health monitoring of thermal power plants[J]. Applied Energy, 2020, 265:114775.
• [9]张珈铭,田静菲,吴哲,等.直接空冷机组凝汽器动态系统参数辨识[J].能源与节能, 2021(11):89-94.ZHANG Jiaming, TIAN Jingfei, WU Zhe, et al. Parameter identification of condenser dynamic system of direct aircooled unit[J]. Energy and Energy Conservation,2021(11):89-94.
• [10] HAFFEJEE R A, LAUBSCHER R. Application of machine learning to develop a real-time air-cooled condenser monitoring platform using thermofluid simulation data[J]. Energy and AI, 2021, 3:100048.
• [11]刘宇航,顾煜炯,郑庆帅,等.基于数据挖掘的直接空冷机组背压预测及优化[J].热力发电, 2023, 52(5):127-135.LIU Yuhang, GU Yujiong, ZHENG Qingshuai, et al.Prediction and optimization of backpressure of direct aircooled unit based on data mining[J]. Thermal Power Generation, 2023, 52(5):127-135.
• [12] RAIDOO R, LAUBSCHER R. Data-driven forecasting with model uncertainty of utility-scale air-cooled condenser performance using ensemble encoder-decoder mixture-density recurrent neural networks[J]. Energy,2022, 238:122030.
• [13] GUYON I, ELISSEEFF A. An introduction to variable and feature selection[J]. Journal of machine learning research, 2003, 3:1157-1182.
• [14] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8):1735-1780.
• [15] POLLACK J B. Recursive distributed representations[J].Artificial Intelligence, 1990, 46(1-2):77-105.
• [16] LECUN Y, BENGIO Y, HINTON G. Deep learning[J].Nature, 2015, 521(7553):436-444.
• [17]高天龙,刘卫亮,张文军.基于门控循环神经网络的光伏阵列故障识别与定位研究[J].热力发电, 2022,51(3):21-28.GAO Tianlong, LIU Weiliang, ZHANG Wenjun. Fault identification and location strategy of photovoltaic power station based on gated recurrent unit[J]. Thermal Power Generation, 2022, 51(3):21-28.
• [18] MINH-THANG L, HIEU P, CHRISTOPHER D. effective approaches to attention-based neural machine translation[C]//Conference on empirical methods in natural language processing, vol.2:Conference on empirical methods in natural language processing(EMNLP 2015), Lisbon, Portugal:Association for Computational Linguistics, 2015:1412-1421.
• [19] WANG D Q, DING F. Hierarchical least squares estimation algorithm for Hammerstein-Wiener systems[J].IEEE Signal Processing Letters, 2012, 19(12):825-828.
• [20] DING F, CHEN T. Hierarchical gradient-based identification of multivariable discrete-time systems[J].Automatica, 2005, 41(2):315-325.
• [21] SHEIKHPOUR R, SARRAM M A, GHARAGHANI S,et al. A survey on semi-supervised feature selection methods[J]. Pattern Recognition, 2017, 64:141-158.
• [22]李汪繁,丁先,方晶剑.基于GWO-RF的凝汽器真空预测方法[J].动力工程学报, 2023, 43(4):436-442.LI Wangfan, DING Xian, FANG Jingjian. Prediction method of condenser vacuum based on GWO-RF[J].Journal of Chinese Society of Power Engineering, 2023,43(4):436-442.
• [23]付红玉,贺红.深度森林在用户购买预测中的应用[J].计算机应用与软件, 2023, 40(1):298-305.FU Hongyu, HE Hong. The application of deep forest in the prediction of user’s purchase behavior[J]. Computer Applications and Software, 2023, 40(1):298-305.
• [24] SRIVASTAVA N, HINTON G, KRIZHEVSKY A, et al.Dropout:a simple way to prevent neural networks from overfitting[J]. The Journal of Machine Learning Research,2014, 15(1):1929-1958.
• [25] KINGMA D P, BA J. Adam:a method for stochastic optimization[J]. Computer Science, 2015, 1412:6980.
• [26] STRU?NIK D. Integration of machine learning to increase steam turbine condenser vacuum and efficiency through gasket resealing and higher heat extraction into the atmosphere[J]. International Journal of Energy Research, 2022, 46(3):3189-3212.