快速准确预测风电机组载荷对机组设计及安全运行具有重要意义。该文通过机组原型测试和数据标定得到风电机组关键部位载荷数据,利用Pearson系数分析多种实测机组状态数据和气象数据的统计量与载荷特性之间的相关性,基于相关性排序确定预测模型输入参数。采用极端随机森林算法建立风电机组载荷预测模型,全面预测机组关键部位极限载荷、平均载荷和等效疲劳载荷。测试结果表明： 预测模型能快速准确预测风电机组叶片、塔顶和塔底载荷特性,平均R2确定系数为0.96,可为机组载荷水平监测和安全运行提供有效支持。
Abstract
Quick and accurate load prediction is crucial to wind turbine design and safe operation. The loads on the wind turbine are acquired by prototype test and data calibration in this study, and the Pearson coefficient method is then used to analyze the correlation relationships of the load characteristics with the statistical data from experimental turbine and meteorological measurements. The input variables for prediction model are determined based on the correlation ranking. The prediction model for wind turbine loads is established using the extremely random forests algorithm to predict the ultimate load, average load and equivalent fatigue load of turbine at crucial positions. The validated results show that the trained model can predict the load characteristics of the blade, tower top, tower bottom quickly and accurately, and the average determination coefficient R2 reaches to 0.96. Therefore, the constructed model can provide effective support for the load monitoring and safe operation of wind turbine.
关键词
风电机组 /
载荷 /
机器学习 /
预测模型 /
实测数据
Key words
wind turbines /
loading /
machine learning /
prediction model /
experimental data
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 付德义, 张晓东, 王瑞明, 等. 特定场址条件下风电机组载荷适应性评估[J]. 太阳能学报, 2021, 42(6): 425-431.
FU D Y, ZHANG X D, WANG R M, et al.Wind turbine load adaptability assessment under specific site conditions[J]. Acta energiae solaris sinica, 2021, 42(6): 425-431.
[2] 董礼, 廖明夫, Martin Kuehn, 等. 风力机等效载荷的评估[J]. 太阳能学报, 2008, 29(12): 1456-1459.
DONG L, LIAO M F, KUEHN M, et al.Estimation of equivalent load for a horizontal axis wind turbine[J]. Acta energiae solaris sinica, 2008, 29(12): 1456-1459.
[3] OBDAM T S, RADEMAKERS L W M M, BRAAM H. Flight leader concept for wind farm load counting: offshore evaluation[J]. Wind engineering, 2010, 34(1): 109-121.
[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 B, YI H Y, WANG X.A model of wind turbine blade root loads based on extreme learning machine[J]. Journal of vibration and shock, 2018, 37(4): 257-262.
[5] 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 S D, XUE Y, MA X J, et al.Prediction of wind turbine key load based on SCADA data[J]. Transactions of the Chinese Society of Agricultural Engineering, 2018, 34(2): 219-225.
[6] 章培成, 孙勇, 王琳. 一种大型风力发电机组载荷预测技术研究[J]. 能源工程, 2020(5): 55-57, 64.
ZHANG P C, SUN Y, WANG L.Study on prediction technology for large wind turbine load[J]. Energy engineering, 2020(5): 55-57, 64.
[7] 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 C, DAI J C, YANG X, et al.Research on blade load identification of large-scale wind turbines based on “stress-load” model[J]. Acta energiae solaris sinica, 2019, 40(5): 1423-1432.
[8] GB/T 37257—2018, 风力发电机组机械载荷测量[S].
GB/T 37257—2018, Wind turbines-measurement of mechanical loads[S].
[9] 付德义, 薛扬, 焦渤, 等. 湍流强度对风电机组疲劳等效载荷的影响[J]. 华北电力大学学报(自然科学版), 2015, 42(1): 45-50.
FU D Y, XUE Y, JIAO B, et al.Effects on the turbulence intensity to wind turbine fatigue equivalent load[J]. Journal of North China Electric Power University (natural science edition), 2015, 42(1): 45-50.
[10] IEC 61400-1, Wind turbines—part 1: design requirements (edition 3)[S].
[11] 李洁明, 祁新娥. 统计学原理[M]. 4版. 上海: 复旦大学出版社, 2007: 1-382.
LI J M, QI X E.General theory of statistics[M]. 4th ed. Shanghai: Fudan University Press, 2007: 1-382.
[12] GEURTS P, ERNST D, WEHENKEL L.Extremely randomized trees[J]. Machine learning, 2006, 63(1): 3-42.
[13] BREIMAN L, FRIEDMAN J H, OLSHEN R A, et al.Classification and regression trees (CART)[J]. Biometrics, 1984, 40(3): 358.
[14] BREIMAN L.Random forests[J]. Machine learning, 2001, 45(1): 5-32.
[15] 周志华. 机器学习[M]. 北京: 清华大学出版社, 2016: 1-425.
ZHOU Z H.Machine learning[M]. Beijing: Tsinghua University Press, 2016: 1-425.
PDF(1991 KB)
