(3)国、内外研究现状和发展动态
人工智能(Artificial Intelligence, AI)技术近年来正处于快速发展阶段,受益于计算机运算能力的提升,AI技术已经可以在许多任务上超越人类[1]。AI技术可以用于模拟物理问题,具有很好的泛化能力和统计推断能力,可为物理建模或实际工程解决方案提供一种替代方法,涵盖了数据分析、预测模型、自动化控制和可再生能源等多个领域。例如:AI模型能够通过分析海洋环境数据来预测精确的海洋波动、风速和其他环境参数,这对于海洋工程的规划和运营至关重要[2-3]。此外,AI技术还在开发自动化控制系统和决策支持系统方面发挥着重要作用,这些系统可以实时监测海洋工程设施的状态,并在必要时提供决策支持。
AI技术在海上风电领域的应用主要体现在动力响应预报方面。动力响应预报在海上风电领域中是十分重要的,而AI和机器学习技术在这过程中发挥着关键作用,它们能够从大量的环境数据中提取出关键特征,进而实现风力机在不同海洋环境下的响应预测。这些技术的应用主要集中在风速和功率预测[4]、风力机性能优化、机械组件监测[5]、故障预测、运维计划的优化[6]、尾流模型的优化[7-11]、多目标预测[12-15]等方面。这种结合使风力机的运行和维护更加智能化,有助于提高风力机的性能、降低事故风险并提高运营效率。目前,使用AI技术对海上风力机动力响应预报的研究主要分为风力机气动载荷预报[16]、海上风力机上部结构动力响应分析[17-20]和系泊张力与平台运动响应[21-25]三个部分。早期学者利用机器学习算法对不同形状的气流漩涡结构进行了识别[26]。Japer等人[27]采用人工神经网络(Artificial Neural Network,ANN),以来流风速、风向和风力机位置作为输入变量,以风电场功率作为输出变量,实现了不同风况和不同风力机位置下风电场的功率预测。Zhang等人[28]基于卷积神经网络(Convolutional Neural Network,CNN),通过输入多个时刻的瞬时三维流域来预测时均化流场。Sun等人[29]同样基于ANN搭建以来流风速、风向和偏航角作为输入的模型,从而预测风力机功率。Optis和Perr-Sauer[30]亦采用ANN搭建模型,以大气来流特征作为输入变量,以风力机功率作为输出,研究了湍流特性和大气稳定性对机器学习方法预测风电场功率的影响。Purohit等人[31]基于ANN搭建以风速、推力系数和湍流强度作为输入的模型,预测了风力机尾流速度场和湍流强度场。部分学者基于ANN建立了尾流预测模型,研究来流风速和湍流强度与风力尾流速度场和湍流场之间的演变关系[32,33]。
相比与实测数据与基于数值仿真的传统动力响应预报方式,使用AI技术可以提高海上风力机安全与稳定性的短期预测能力,并仰仗其极快的输出速率,使动力响应具备了实时预报的潜力。但需要指出的是,尽管AI技术在动力响应预报中表现出潜力,但其仍然依赖于基础数据库的建立,用以代理模型的训练,同时需要经过海上风力机试验或实测数据的验证,以确保代理模型在实际应用中的可靠性和准确性。
参考文献
[1] DONG H, DING Z H,ZHANG S H. Deep Reinforcement Learning[M]. Singapore: Springer Singapore, 2020.
[2] HOU Y L, WANG Q W. Big data and artificial intelligence application in energy field: a bibliometric analysis [J].Environmental Science and Pollution Research, 2023(6):13960-13973..
[3] MA Y. Machine learning in ocean applications: wave prediction for advanced controls of renewable energy and modeling nonlinear viscous hydrodynamics[D]. Cambridge: Massachusetts Institute of Technology, 2020.
[4] ATEŞ K T. Estimation of short-term power of wind turbines using artificial neural network( ANN) and swarm intelligence [J]. Sustainability, 2023(18): 13572.
[5] FARRAR N O, ALI M H, DASGUPTA D. Artificial intelligence and machine learning in grid connected wind turbine control systems:a comprehensive review[J]. Energies, 2023( 3): 1530.
[6] VIVES J, PALACI J, HEART J. Framework for bidirectional knowledge-based maintenance of wind turbines[J]. Computational Intelligence and Neuroscience, 2022, 2022 : 1-6.
[7] [5] M. Fahim, V. Sharma, T. -V. Cao, B. Canberk and T. Q. Duong, "Machine Learning-Based Digital Twin for Predictive Modeling in Wind Turbines," in IEEE Access, vol. 10, pp. 14184-14194, 2022.
[8] Y. Huang, S. Lin and X. Zhao, "Multi-Agent Reinforcement Learning Control of a Hydrostatic Wind Turbine-Based Farm," in IEEE Transactions on Sustainable Energy, vol. 14, no. 4, pp. 2406-2416, Oct. 2023.
[9] Zhiwen Deng, Chang Xu, Xingxing Han, Zhe Cheng, Feifei Xue, Decentralized yaw optimization for maximizing wind farm production based on deep reinforcement learning, Energy Conversion and Management, Volume 286, 2023, 117031, ISSN 0196-8904.
[10] Kun Yang, Xiaowei Deng, Layout optimization for renovation of operational offshore wind farm based on machine learning wake model, Journal of Wind Engineering and Industrial Aerodynamics, Volume 232, 2023, 105280, ISSN 0167-6105.
[11] Rui Li, Jincheng Zhang, Xiaowei Zhao, Dynamic wind farm wake modeling based on a Bilateral Convolutional Neural Network and high-fidelity LES data, Energy, Volume 258, 2022, 124845.
[12] Xiuxing Yin, Xiaowei Zhao, Big data driven multi-objective predictions for offshore wind farm based on machine learning algorithms, Energy, Volume 186, 2019, 115704.
[13] Xiuxing Yin, Wencan Zhang, Zhansi Jiang, Li Pan, Data-driven multi-objective predictive control of offshore wind farm based on evolutionary optimization, Renewable Energy, Volume 160, 2020, Pages 974-986.
[14] Xiuxing Yin, Xiaowei Zhao, Jin Lin, Aris Karcanias, Reliability aware multi-objective predictive control for wind farm based on machine learning and heuristic optimizations, Energy, Volume 202, 2020, 117739.
[15] Ruiyang He, Hongxing Yang, Shilin Sun, Lin Lu, Haiying Sun, Xiaoxia Gao, A machine learning-based fatigue loads and power prediction method for wind turbines under yaw control, Applied Energy, Volume 326, 2022, 120013.
[16] JIANG X, DAY S, CLELLAND D, et al. Analysis and realtime prediction of the full-scale thrust for floating wind turbine based on artificial intelligence[J]. Ocean Engineering, 2019, 175: 207-216.
[17] PAN L, WANG X D. Variable pitch control on direct-driven PMSG for offshore wind turbine using Repetitive-TS fuzzy 22 PID control[J]. Renewable Energy, 2020, 159: 221-237.
[18] SIERRA-GARCIA J E, SANTOS M. Deep learning and fuzzy logic to implement a hybrid wind turbine pitch control[J]. Neural Computing and Applications, 2022 (13): 10503-10517.
[19] WANG Z M, QIAO D S, TANG G Q, et al. An identification method of floating wind turbine tower responses using deep learning technology in the monitoring system[J]. Ocean Engineering, 2022, 261: 112105.
[20] SIMPSON T, DERVILIS N, COUTURIER P, et al. Reduced order modeling of non-linear monopile dynamics via an AE-LSTM scheme[J]. Frontiers in Energy Research, 2023, 11: 1128201.
[21] 张德庆, 王超, 杜君峰. 基于人工神经网络算法的深海浮式系统动力响应预报方法[J]. 中国造船, 2021 (1): 123-132.
[22] WANG Z M, QIAO D S, YAN J, et al. A new approach to predict dynamic mooring tension using LSTM neural network based on responses of floating structure[J]. Ocean Engineering, 2022, 249: 110905.
[23] QIAO D S, LI P, MA G, et al. Realtime prediction of dynamic mooring lines responses with LSTM neural network model [J]. Ocean Engineering, 2021, 219: 108368.
[24] MAO Y X, WANG T Q, DUAN M L. A DNN-based approach to predict dynamic mooring tensions for semi-submersible platform under a mooring line failure condition[J]. Ocean Engineering, 2022, 266: 112767.
[25] SHI W, HU L H, LIN Z B, et al. Short-term motion prediction of floating offshore wind turbine based on muti-input LSTM neural network[J]. Ocean Engineering, 2023, 280: 114558.
[26] Spencer A J M, Rivlin R S. Isotropic integrity bases for vectors and second-order tensors - Part I[J]. Archive for Rational Mechanics and Analysis, 1962, 9(1): 45–63.
[27] Japar F , Mathew S , Narayanaswamy B , et al. Estimating the wake losses in large wind farms: A machine learning approach[C]// Innovative Smart Grid Technologies Conference. IEEE, 2014: 1–5.
[28] Zhang Z, Santoni C, Herges T, et al. Time-Averaged Wind Turbine Wake Flow Field Prediction Using Autoencoder Convolutional Neural Networks[J]. Energies, 2021, 15(1): 1-20.
[29] Sun H, Qiu C, Lu L, et al. Wind turbine power modelling and optimization using artificial neural network with wind field experimental data[J]. Applied Energy, 2020, 280(12): 1-12.
[30] Optis M, Perr-Sauer J. The importance of atmospheric turbulence and stability in machine-learning models of wind farm power production[J]. Renewable and Sustainable Energy Reviews, 2019, 112(9): 27-41.
[31] Purohit S, Ng E Y K, Syed Ahmed Kabir I F. Evaluation of three potential machine learning algorithms for predicting the velocity and turbulence intensity of a wind turbine wake[J]. Renewable Energy, 2022, 184(1): 405–420.
[32] Ti Z, Deng X W, Yang H. Wake modeling of wind turbines using machine learning[J]. Applied Energy, 2020, 257(1): 1-17.
[33] Ti Z, Deng X W, Zhang M. Artificial Neural Networks based wake model for power prediction of wind farm[J]. Renewable Energy, 2021, 172(7): 618-631.