融合光纤传感数据与门控循环神经网络的风电机组载荷预测方法

doi:10.3901/JME.2025.08.272

机械工程学报 ›› 2025, Vol. 61 ›› Issue (8): 272-282.doi: 10.3901/JME.2025.08.272

• 可再生能源与工程热物理 • 上一篇

扫码分享

融合光纤传感数据与门控循环神经网络的风电机组载荷预测方法

肖钊¹, 曹智辉², 邓杰文¹, 段书用³, 赵前程¹, 戴巨川¹, 陶洁²

1. 湖南科技大学机电工程学院湘潭 411201;
2. 湖南科技大学计算机科学与工程学院湘潭 411201;
3. 智能配用电装备与系统全国重点实验室(河北工业大学) 天津 300401

收稿日期:2024-06-21 修回日期:2024-10-11 发布日期:2025-05-10
作者简介:肖钊，男，1987年出生，博士，副教授。主要研究方向为风电大数据挖掘和结构可靠性优化设计。E-mail：xnxzh501@hnust.edu.cn;陶洁(通信作者)，女，1980年出生，博士，副教授。主要研究方向为机器学习、数据挖掘、机械故障诊断等。E-mail：caroltao@hnust.edu.cn
基金资助:
国家重点研发计划(2022YFF0608700)、湖南省杰出青年基金(2024JJ2031)、湖南省自然科学基金(2023JJ30265)和湖南省科技创新计划(2023RC3174)资助项目。

Wind Turbine Load Prediction Method Combining Fiber Bragg Grating Sensor Data and Gated Recurrent Units Neural Network

XIAO Zhao¹, CAO Zhihui², DENG Jiewen¹, DUAN Shuyong³, ZHAO Qiancheng¹, DAI Juchuan¹, TAO Jie²

1. School of Mechanical and Electrical Engineering, Hunan University of Science and Technology, Xiangtan 411201;
2. School of Computer Science and Engineering, Hunan University of Science and Technology, Xiangtan 411201;
3. State Key Laboratory of Intelligent Power Distribution Equipment and System, Hebei University of Technology, Tianjin 300401

Received:2024-06-21 Revised:2024-10-11 Published:2025-05-10

摘要/Abstract

摘要： 风电机组服役环境恶劣，对运行载荷的监测和分析是提升设备可靠性和发电性能的关键因素，由于载荷监测设备成本高昂，现役风电机组并未大范围安装载荷传感器，对风机载荷进行实时监测。为此，提出一种基于门控循环神经网络(Gated recurrent unit neural network，GRU)的风机载荷预测方法，利用少量载荷测试数据和数据采集与监视控制系统(Supervisory control and data acquisition，SCADA)数据进行载荷预测。该方法采用机理知识和皮尔逊相关性分析确定影响风电机组运行载荷关键参数，然后选取SCADA中相关的实时数据作为模型输入，通过门控循环神经网络实现风机载荷的实时预测。试验中对极端运行工况下的载荷进行预测，并将GRU模型与MLP、LSTM、RNN模型进行了对比，结果表明GRU模型在预测精度及误差最小化方面表现更好。最后，通过多参数的消融试验，证明GRU模型具有较强的鲁棒性，可以利用现有SCADA数据实现对大规模风电机组载荷进行预测，为进一步风电机组结构状态评估和寿命预测提供支撑。

关键词: 风电机组载荷预测, 光纤光栅传感器, 门控循环神经网络, 状态评估

Abstract: Wind turbines operate in harsh environments, and monitoring and analyzing operating loads are key factors in improving equipment reliability and power generation performance. Due to the high cost of load monitoring equipment, load sensors are not widely installed on existing wind turbines for real-time monitoring. Therefore, a wind turbine load prediction method based on gated recurrent units neural network(GRU) is proposed, which uses a small amount of load test data and supervisory control and data acquisition(SCADA) data for load prediction. This method uses mechanistic knowledge and pearson correlation analysis to determine the key parameters affecting the operating load of wind turbines, then selects relevant real-time data in SCADA as model input, and realizes the real-time prediction of wind turbine load through a gated recurrent neural network. In the experiment, the load under extreme operating conditions is predicted, and the GRU model is compared with the MLP, LSTM, and RNN models. The results show that the GRU model performed better in prediction accuracy and error minimization. Finally, through multi-parameter ablation experiments, it is proved that the GRU model has strong robustness and can use existing SCADA data to predict the load of large-scale wind turbines. And this method can provide support for further structural status assessment and life prediction of wind turbines.

Key words: load prediction of wind turbines, fiber bragg grating sensor, gated recurrent neural network, state assessment

中图分类号:

TM315

肖钊, 曹智辉, 邓杰文, 段书用, 赵前程, 戴巨川, 陶洁. 融合光纤传感数据与门控循环神经网络的风电机组载荷预测方法[J]. 机械工程学报, 2025, 61(8): 272-282.

XIAO Zhao, CAO Zhihui, DENG Jiewen, DUAN Shuyong, ZHAO Qiancheng, DAI Juchuan, TAO Jie. Wind Turbine Load Prediction Method Combining Fiber Bragg Grating Sensor Data and Gated Recurrent Units Neural Network[J]. Journal of Mechanical Engineering, 2025, 61(8): 272-282.

参考文献

[1] BOŠNJAKOVIĆ M, KATINIĆ M, SANTA R，et al. Wind turbine technology trends[J]. Applied Sciences，2022，12(17)：8653.
[2] REN Zhengru，VERMA A S，LI Ye，et al. Offshore wind turbine operations and maintenance：A state-of-the-art review[J]. Renewable and Sustainable Energy Reviews，2021，144：110886.
[3] LIAO Ding，ZHU Shunpeng，CORREIA J，et al. Fatigue reliability of wind turbines：Historical perspectives，recent developments and future prospects[J]. Renewable Energy，2022，200：724-742.
[4] 秦斌，易怀洋，王欣. 基于极限学习机的风电机组叶根载荷辨识建模[J]. 振动与冲击，2018，37(4)：257-262. QIN Bin，YI Huaiyang，WANG Xin. Identification and modeling of wind turbine blade root load based on extreme learning machine[J]. Vibration and Impact，2018，37(4)：257-262.
[5] 王思清，滕伟，王罗，等. 水平轴风电机组叶片摆振与挥舞机理建模[J]. 机械工程学报，2024，60(3)：131-142. WANG Siqing，TENG Wei，WANG Luo，et al. Modeling of blade oscillation and flapping mechanism in horizontal axis wind turbines[J]. Journal of Mechanical Engineering，2024，60(3)：131-142.
[6] WANG Zedong，LU Ziqi，YI Wenwu，et al. A study of nonlinear aeroelastic response of a long flexible blade for the horizontal axis wind turbine[J]. Ocean Engineering，2023，279：113660.
[7] CHEN Chaohe，MA Yuan，FAN Tianhui. Review of model experimental methods focusing on aerodynamic simulation of floating offshore wind turbines[J]. Renewable and Sustainable Energy Reviews，2022，157：112036.
[8] AMIRI M M，SHADMAN M，ESTEFEN S F. A review of physical and numerical modeling techniques for horizontal-axis wind turbine wakes[J]. Renewable and Sustainable Energy Reviews，2024，193：114279.
[9] 李密密，戴巨川，陈换过，等. 基于光纤传感的风电机组叶片应变传递特性研究[J]. 太阳能学报，2022，43(7)：375-384. LI Mimi，DAI Juchuan，CHEN Huanguo，et al. Research on strain transfer characteristics of wind turbine blades based on fiber optic sensing [J]. Journal of Solar Energy，2022，43(7)：375-384.
[10] GUTIERREZ S U，FERNÁNDEZ S A，POLINDER H，et al. Input torque measurements for wind turbine gearboxes using fiber-optic strain sensors[J]. Wind Energy Science，2022，7(2)：505-521.
[11] 王超，戴巨川，杨鑫，等. 基于“应变-载荷”模型的大型风电机组叶片载荷识别研究[J]. 太阳能学报，2019，40(5)：1423-1432. WANG Chao，DAI Juchuan，YANG Xin，et al. Research on load identification of large wind turbine blades based on the “strain load” model[J]. Journal of Solar Energy，2019，40(5)：1423-1432.
[12] ZHU Fan，ZHANG Meng，MA Fuxuan，et al. Identification of wind load exerted on the jacket wind turbines from optimally placed strain gauges using C-optimal design and mathematical model reduction[J]. Journal of Marine Science and Engineering，2024，12(4)：563.
[13] WEN Binrong，TIAN Xinliang，JIANG Zhihao，et al. Monitoring blade loads for a floating wind turbine in wave basin model tests using fiber bragg grating sensors：A feasibility study[J]. Marine Structures，2020，71：102729.
[14] KAEWNIAM P，CAO M，ALKAYEM N F，et al. Recent advances in damage detection of wind turbine blades：A state-of-the-art review[J]. Renewable and Sustainable Energy Reviews，2022，167：112723.
[15] 李新华，梁浩，徐伟弘，等. 常用分布式光纤传感器性能比较[J]. 光通信技术，2007(5)：14-18. LI Xinhua，LIANG Hao，XU Weihong，et al. Performance comparison of commonly used distributed fiber optic sensors[J]. Optical Communication Technology，2007(5)：14-18.
[16] DAI Wei，LI Dongsheng，JIANG Tao，et al. Load identification of a 2.5 MW wind turbine tower using Kalman filtering techniques and BDS data[J]. Engineering Structures，2023，281：115763.
[17] NOPPE N，WEIJTJENS W，DEVRIENDT C. Modeling of quasi-static thrust load of wind turbines based on 1 s SCADA data[J]. Wind Energy Science，2018，3(1)：139-147.
[18] 周士栋，薛扬，马晓晶，等. 基于SCADA数据的风电机组关键载荷预测[J]. 农业工程学报，2018，34(2)：219-225. ZHOU Shidong，XUE Yang，MA Xiaojing，et al. Wind turbine critical load prediction based on SCADA data[J] Journal of Agricultural Engineering，2018，34(2)：219-225.
[19] 牟哲岳，孙勇，王瑞良，等. 基于实测数据和机器学习的风电机组载荷预测模型[J]. 太阳能学报，2023，44(10)：414-419. MOU Zheyue，SUN Yong，WANG Ruiliang，et al. Wind turbine load prediction model based on measured data and machine learning[J]. Journal of Solar Energy，2023，44(10)：414-419.
[20] DAI Juchuan，LI Mimi，CHEN Huanguo，et al. Progress and challenges on blade load research of large-scale wind turbines[J]. Renewable Energy，2022，196：482-496.
[21] SHEWALKAR A，NYAVANANDI D，LUDWIG S A. Performance evaluation of deep neural networks applied to speech recognition：RNN，LSTM and GRU[J]. Journal of Artificial Intelligence and Soft Computing Research，2019，9(4)：235-245.
[22] YAMAK P T，YUJIAN L，GADOSEY P K. A comparison between ARIMA，LSTM，and GRU for time series forecasting[C]//Proceedings of the 20192nd International Conference on Algorithms，Computing and Artificial Intelligence. 2019：49-55.
[23] LIASHCHYNSKYI P. Grid search，random search，genetic algorithm：A big comparison for NAS[J]. arXiv Preprint arXiv，2019：1912.06059.
[24] POPESCU M C，BALAS V E，PERESCU-POPESCU L，et al. Multilayer perceptron and neural networks[J]. WSEAS Transactions on Circuits and Systems，2009，8(7)：579-588.
[25] YU Yong，SI Xiaosheng，HU Changhua ，et al. A review of recurrent neural networks：LSTM cells and network architectures[J]. Neural Computation，2019，31(7)：1235-1270.
[26] SHERSTINSKY A. Fundamentals of recurrent neural network and long short-term memory network[J]. Physica D：Nonlinear Phenomena，2020，404：132306.

融合光纤传感数据与门控循环神经网络的风电机组载荷预测方法

Wind Turbine Load Prediction Method Combining Fiber Bragg Grating Sensor Data and Gated Recurrent Units Neural Network

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 6

编辑推荐

Metrics

本文评价

[1]	蒋翔宇, 冯毅雄, 洪兆溪, 胡炳涛, 司恒远, 谭建荣. 基于多级数据融合的核电站运行功能健康状态评估[J]. 机械工程学报, 2024, 60(4): 200-211.
[2]	刘思远, 艾超, 郁春嵩, 张伟哲, 陈文婷, 康伟. 基于深度信念网络多参数智能融合的滑靴副性能退化状态评估方法[J]. 机械工程学报, 2024, 60(4): 178-188.
[3]	郭锐, 张印浩, 牛雯雯, 骆雄帅, 蔡伟, 王建伟, 王岳峰, 赵静一. 基于GLCT及CPA-SVM的变转速齿轮泵健康状态分类研究[J]. 机械工程学报, 2023, 59(14): 310-319.
[4]	余浩, 黄彦钦, 尹钧毅, 孟国栋, 成永红. 基于介电响应特性的油浸式电力变压器绝缘状态评估研究进展[J]. 电气工程学报, 2020, 15(4): 1-17.
[5]	朱万旭, 覃荷瑛, 李居泽, 欧进萍. 基于锚固区外置光纤光栅传感器的FAST工程拉索索力监测研究[J]. 机械工程学报, 2017, 53(17): 23-30.
[6]	朱小军;陈雪峰;翟智;田绍华;何正嘉. 复合材料风电叶片细观失效准则及其损伤演化[J]. , 2014, 50(11): 17-22.