为评估风电机组的健康状态,提出一种基于CNN-BiGRU的风电机组健康状态评估方法。首先,采用四分位法对风电机组数据采集与监视控制（SCADA）系统中存在的异常数据进行剔除,并采用最大相关最小冗余（mRMR）算法实现与功率相关特征的选取。然后,以预测风电机组的输出功率为目标,构建基于CNN-BiGRU的功率预测模型,采用卷积神经网络（CNN）实现对输入高维数据特征的有效提取,同时采用麻雀搜索算法（SSA）对双向门控循环单元（BiGRU）网络参数进行优化;以风电机组处于健康状态下的功率预测残差建立基准分布模型,计算实时预测结果的残差到基准分布模型的马氏距离（MD）,根据该马氏距离构建风电机组健康指标,实现风电机组的健康状态评估。
Abstract
To evaluate the health status of wind turbines, this paper proposes a CNN-BiGRU-based health evaluation method. Firstly, the quartile method is used to eliminate the abnormal data in the wind turbine supervisory control and data acquisition （SCADA） system, and the maximum relevance minimum redundancy （mRMR） algorithm is applied to select power-related features. Then,a CNN-BiGRU-based power prediction model is constructed to predict the wind turbine's output power.The convolutional neural network （CNN） extracts high-dimensional input features effectively, while the sparrow search algorithm （SSA） optimizes the parameters of the bidirectional gated recurrent unit （BiGRU） network. A benchmark distribution model is established based on the power prediction residuals of wind turbines in a healthy state, calculate the Mahalanobis distance （MD） from the residuals of real-time prediction results to the benchmark distribution model, and construct wind turbine health indicators based on this MD to evaluate the health status of wind turbine.
关键词
风电机组 /
风电机组数据采集与监视控制（SCADA） /
功率预测 /
健康状态评估 /
卷积神经网络 /
双向门控循环单元
Key words
wind turbines /
supervisory control and data acquisition,（SCADA） /
power prediction /
health status assessment /
convolutional neural network /
BIGRU
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 许帅, 霍志红, 许昌, 等. 限功率控制下风电机组叶片疲劳损伤研究[J]. 太阳能学报, 2020, 41(1): 282-289.
XU S, HUO Z H, XU C, et al.Study on fatigue damage of wind turbine blade under power curtailment control condition[J]. Acta energiae solaris sinica, 2020, 41(1): 282-289.
[2] REN C, XING Y H.An efficient active learning Kriging approach for expected fatigue damage assessment applied to wind turbine structures[J]. Ocean engineering, 2024, 305: 118034.
[3] 李东东, 刘宇航, 赵阳, 等. 基于改进生成对抗网络的风机行星齿轮箱故障诊断方法[J]. 中国电机工程学报, 2021, 41(21): 7496-7507.
LI D D, LIU Y H, ZHAO Y, et al.Fault diagnosis method of wind turbine planetary gearbox based on improved generative adversarial network[J]. Proceedings of the CSEE, 2021, 41(21): 7496-7507.
[4] LIU Z P, YANG B Y, WANG X F, et al.Acoustic emission analysis for wind turbine blade bearing fault detection under time-varying low-speed and heavy blade load conditions[J]. IEEE transactions on industry applications, 2021, 57(3): 2791-2800.
[5] LIU Z P, WANG X F, ZHANG L.Fault diagnosis of industrial wind turbine blade bearing using acoustic emission analysis[J]. IEEE transactions on instrumentation and measurement, 2020, 69(9): 6630-6639.
[6] 金晓航, 秦治伟, 郭远晶, 等. 基于改进劣化度模型的风电机组日常运行状态评估[J]. 太阳能学报, 2023, 44(1): 239-246.
JIN X H, QIN Z W, GUO Y J, et al.Evaluation of wind turbines daily operation conditions based on improved degradation model[J]. Acta energiae solaris sinica, 2023, 44(1): 239-246.
[7] ZHOU Z H, ZHU J P, FU Z X, et al.Offshore wind turbine condition assessment method based on GRU and improved fuzzy comprehensive evaluation[C]//2022 IEEE 5th International Electrical and Energy Conference (CIEEC). Nangjing, China, 2022: 3728-3735.
[8] XIANG L, YANG X, HU A J, et al.Condition monitoring and anomaly detection of wind turbine based on cascaded and bidirectional deep learning networks[J]. Applied energy, 2022, 305: 117925.
[9] CHEN W H, ZHOU H T, CHENG L S, et al.Condition monitoring of wind turbine using novel deep learning method and dynamic kernel principal components Mahalanobis distance[J]. Engineering applications of artificial intelligence, 2023, 125: 106757.
[10] DAO P B.Condition monitoring and fault diagnosis of wind turbines based on structural break detection in SCADA data[J]. Renewable energy, 2022, 185: 641-654.
[11] YAO Q T, BING H K, ZHU G P, et al.A novel stochastic process diffusion model for wind turbines condition monitoring and fault identification with multi-parameter information fusion[J]. Mechanical systems and signal processing, 2024, 214: 111397.
[12] 李昕燃, 靳伍银. 基于改进麻雀算法优化支持向量机的滚动轴承故障诊断研究[J]. 振动与冲击, 2023, 42(6): 106-114.
LI X R, JIN W Y.Fault diagnosis of rolling bearings based on ISSA-SVM[J]. Journal of vibration and shock, 2023, 42(6): 106-114.
[13] 姚程文, 杨苹, 刘泽健. 基于CNN-GRU混合神经网络的负荷预测方法[J]. 电网技术, 2020, 44(9): 3416-3423.
YAO C W, YANG P, LIU Z J.Load forecasting method based on CNN-GRU hybrid neural network[J]. Power system technology, 2020, 44(9): 3416-3423.
[14] 刘赟. ReLU激活函数下卷积神经网络的不同类型噪声增益研究[D]. 南京: 南京邮电大学, 2023.
LIU Y.Study on different types of noise benefit of convolutional neural networks under ReLU activation function[D]. Nanjing: Nanjing University of Posts and Telecommunications, 2023.
[15] 王长江, 张千龙, 姜涛, 等. 基于图卷积网络和双向门控循环单元的电力系统主导失稳模式辨识[J]. 中国电机工程学报, 2025, 45(16): 6326-6340.
WANG C J, ZHANG Q L, JIANG T, et al.Power system dominant instability mode identification based on graph convolutional networks and bidirectional gated recurrent units[J]. Proceedings of the CSEE, 2025, 45(16): 6326-6340.
[16] 李道全, 刘旭寅, 刘嘉宇, 等. 结合Transformer的双向GRU入侵检测研究[J]. 计算机工程与应用, 2025, 61(10): 299-307.
LI D Q, LIU X Y, LIU J Y, et al.Intrusion detection research combining Transformer and bidirectional GRU[J]. Computer engineering and applications, 2025, 61(10): 299-307.
[17] 王朋鹤. 基于双向门控循环神经网络的风电机组状态检测[D]. 北京: 华北电力大学, 2021.
WANG P H.State Detection of wind turbine based on bidirectional gated recurrent unit neural network[D]. Beijing: North China Electric Power University, 2021.
[18] IHIANLE I K, MACHADO P, OWA K, et al.Minimising redundancy, maximising relevance: HRV feature selection for stress classification[J]. Expert systems with applications, 2024, 239: 122490.
[19] 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 status assessment of wind turbine based on gated recurrent unit neural network[D]. Tianjin: Hebei University of Technology, 2021.
[20] YANG J J, DELPHA C.A new reconstruction-based method using local Mahalanobis distance for incipient fault isolation and amplitude estimation[J]. Mechanical systems and signal processing, 2023, 205: 110803.
PDF(20291 KB)
