为了科学合理地反映风电机组齿轮箱运行状态,提出一种基于Vine-Copula模型和双向长短期记忆（BiLSTM）算法的风电机组齿轮箱健康状态评估模型。首先,通过Vine-Copula模型分析数据采集与监视控制（SCADA）系统中各相关状态参数之间的耦合特性,然后利用BiLSTM算法构建健康状态下的标准残差,用于评估风电机组齿轮箱的健康状况。最后,使用实时数据计算残差值并与健康状态下的标准残差值进行比较,利用马氏距离来度量两者之间的差异,并结合健康指数对风电机组齿轮箱的状态等级划分4个等级（优秀、正常、注意和恶劣）。结果表明：针对某风电场发生故障时不同工况的实际数据进行验证,对于不同工况下风电机组齿轮箱油温超温状态,该模型可提前90和1186 min进行故障预警,实现对风电机组齿轮箱运行健康状态的评估。
Abstract
To accurately and scientifically assess the operational condition of wind turbine gearbox, a health status assessment model based on the Vine-Copula model and Bi-directional Long Short-Term Memory （BiLSTM） algorithm is proposed. Firstly, the coupling relationships between various state parameters in the Supervisory Control and Data Acquisition （SCADA） system are analyzed using the Vine-Copula model. Then, the BiLSTM algorithm is employed to construct standard residuals under healthy conditions, which are used to assess the gearbox's health status. Finally, real-time data are used to compute the residuals, which are compared with the standard residuals under healthy conditions. Mahalanobis distance is utilized to measure the difference between the covnputed residual and standard residual, and a health index is incorporated to categorize the gearbox’s health status into four levels: excellent, normal, attentive and poor. The results demonstrate that the model can provide early fault warnings for gearbox oil temperature overheating under various operating conditions, with lead times of 90 minutes and 1186 minutes, respectively. This model effectively enables the assessment of the operational health status of the wind turbine gearbox.
关键词
风电机组 /
状态评估 /
SCADA系统 /
预测分析 /
神经网络
Key words
wind turbines /
status assessment /
SCADA system /
predictive analysis /
neural network
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 杨锡运, 米尔扎提·买合木提, 刘思渠, 等. 基于Vine-Copula贝叶斯网络模型的风机高温降容状态评估方法[J]. 中国电机工程学报, 2020, 40(11): 3583-3592.
YANG X Y, MAIHEMUTI MIERZHATI, LIU S Q, et al.State assessment method of capacity reduction at high temperature for wind turbine based on Vine-Copula Bayesian network model[J]. Proceedings of the CSEE, 2020, 40(11): 3583-3592.
[2] 魏书荣, 张鑫, 符杨, 等. 基于GRA-LSTM-Stacking模型的海上双馈风力发电机早期故障预警与诊断[J]. 中国电机工程学报, 2021, 41(7): 2373-2383.
WEI S R, ZHANG X, FU Y, et al.Early fault warning and diagnosis of offshore wind DFIG based on GRA-LSTM-Stacking model[J]. Proceedings of the CSEE, 2021, 41(7): 2373-2383.
[3] 曾天生, 刘航, 陈汉斯, 等. 基于多特征信息融合的风电机组整机性能评估[J]. 计算机集成制造系统, 2022, 28(4): 1052-1061.
ZENG T S, LIU H, CHEN H S, et al.Performance assessment of wind turbine based on multi-feature information fusion[J]. Computer integrated manufacturing systems, 2022, 28(4): 1052-1061.
[4] 闫海波, 李明明. 基于Vine Copula的投资组合风险度量和相关结构研究[J]. 赤峰学院学报(自然科学版), 2022, 38(5): 40-44.
YAN H B, LI M M.Research on portfolio risk measurement and related structure based on Vine Copula[J]. Journal of Chifeng University(natural science edition), 2022, 38(5): 40-44.
[5] 梁州, 李昊, 林宇. R-Vine Copula、极值理论与股票市场组合风险测度[J]. 南方金融, 2018(7): 28-38.
LIANG Z, LI H, LIN Y.R-Vine Copula, extreme value theory and risk measurement of stock portfolio[J]. South China finance, 2018(7): 28-38.
[6] XIAO Q, ZHOU S W, XIAO H G.Probabilistic optimal power flow analysis incorporating correlated wind sources[J]. International transactions on electrical energy systems, 2020, 30(8): e12441.
[7] YANG X Y, YE T Z, FAN L W.Recognition of blade icing state based on vine-Copula model and LSTM-Autoencoder algorithm[J]. 2021, 13(2): 023312.
[8] ZHAN J, WANG R L, YI L Z, et al.Health assessment methods for wind turbines based on power prediction and mahalanobis distance[J]. International journal of pattern recognition and artificial intelligence, 2019, 33(2): 1951001.
[9] 景彤梅, 齐咏生, 刘利强, 等. 基于KECA-GRNN的风电机组齿轮箱状态监测与健康评估[J]. 太阳能学报, 2021, 42(6): 400-408.
JING T M, QI Y S, LIU L Q, et al.Condition monitoring and health assessment of wind turbine gearbox based on KECA-GRNN[J]. Acta energiae solaris sinica, 2021, 42(6): 400-408.
[10] 何群, 李晔阳, 江国乾, 等. 基于条件卷积自编码高斯混合模型的风电齿轮箱健康评估[J]. 太阳能学报, 2023, 44(12): 214-220.
HE Q, LI Y Y, JIANG G Q, et al.Health assessment of wind turbine gearbox based on conditional convolution autoencoding Gaussian mixture model[J]. Acta energiae solaris sinica, 2023, 44(12): 214-220.
[11] 尹诗, 侯国莲, 迟岩, 等. 风电机组发电机前轴承健康度预测方法及实现[J]. 系统仿真学报, 2021, 33(6): 1323-1333.
YIN S, HOU G L, CHI Y, et al.Prediction method for health degree of front bearing of wind turbine generator and implementation[J]. Journal of system simulation, 2021, 33(6): 1323-1333.
[12] 宋仁旺, 张岩, 石慧. 基于Copula函数的齿轮箱剩余寿命预测方法[J]. 系统工程理论与实践, 2020, 40(9): 2466-2474.
SONG R W, ZHANG Y, SHI H.Prediction method for the remaining useful life of gearbox based on Copula function[J]. Systems engineering-theory & practice, 2020, 40(9): 2466-2474.
[13] 王同. 基于Vine Copula的行业内股票相依结构研究[D]. 北京:北方工业大学,2022.
WANG T.Research on the dependence structure of stocks in the industry based on Vine Copula[D]. Beijing: North China University of Technology, 2022.
[14] 韦颖璐. 基于pair-Copula熵的相关性度量[D]. 苏州: 苏州大学, 2021.
WEI Y L.Correlation measurement based on pair-Copula entropy[D]. Suzhou: Soochow University, 2021.
[15] 蒋楠, 陈维荣, 戴朝华, 等. 一种多维风电场相关性的优化建模方法[J]. 可再生能源, 2017, 35(9): 1359-1367.
JIANG N, CHEN W R, DAI C H, et al.An optimized modeling method for multidimensional wind farms correlation[J]. Renewable energy resources, 2017, 35(9): 1359-1367.
[16] WEI J W, ZHOU L.Model selection using modified AIC and BIC in joint modeling of paired functional data[J]. Statistics & probability letters, 2010, 80(23/24): 1918-1924.
[17] 王蕾. C-Vine Copula在复杂数据上的应用研究[D]. 沈阳: 沈阳师范大学, 2021.
WANG L.Research on the application of C-Vine Copula in complex data[D]. Shenyang: Shenyang Normal University, 2021.
[18] 杨茂, 马剑. 基于Copula理论的风电功率缺失数据补齐方法研究[J]. 电测与仪表, 2018, 55(3): 13-19.
YANG M, MA J.Study on wind power data completion method based on Copula theory[J]. Electrical measurement & instrumentation, 2018, 55(3): 13-19.
[19] LIANG T, MENG Z C, XIE G F, et al.Multi-running state health assessment of wind turbines drive system based on BiLSTM and GMM[J]. IEEE access, 2020, 8: 143042-143054.
[20] 柳青秀, 马红占, 褚学宁, 等. 基于长短时记忆: 自编码神经网络的风电机组性能评估及异常检测[J]. 计算机集成制造系统, 2019, 25(12): 3209-3219.
LIU Q X, MA H Z, CHU X N, et al.Performance assessment and anomaly detection of wind turbine based on long short time memory-auto encoder neural network[J]. Computer integrated manufacturing systems, 2019, 25(12): 3209-3219.
[21] 付雪娇. 风电机组状态参数预测与运行状态评估方法研究[D]. 沈阳: 沈阳工业大学, 2022.
FU X J.Research on state parameter prediction and operation state evaluation method of wind turbine[D]. Shenyang: Shenyang University of Technology, 2022.
[22] 谢高锋. 基于门控循环神经网络的风电机组健康状态评估[D]. 天津: 河北工业大学, 2021.
XIE G F.Health evaluation of wind turbine based on gated cyclic neural network[D]. Tianjin: Hebei University of Technology, 2021.
[23] 李振恩, 张新燕, 胡威, 等. 基于健康指数的风电机组高速轴轴承状态评估与预测[J]. 太阳能学报, 2021, 42(10): 290-297.
LI Z E, ZHANG X Y, HU W, et al.State accessment and prediction of wind turbine high speed shaft bearing based on health index[J]. Acta energiae solaris sinica, 2021, 42(10): 290-297.
[24] 赵祖龙. 基于大数据分析的风电机组健康状态智能评估及诊断探索[J]. 中国管理信息化, 2021, 24(8): 116-117.
ZHAO Z L.Intelligent assessment and diagnosis of wind turbine health based on big data analysis[J]. China management informationization, 2021, 24(8): 116-117.
基金
辽宁省教育厅资助项目（LQGD2020016）
PDF(2293 KB)
