为了提高风电机组变桨系统故障诊断的准确性,提出一种基于批标准化的堆叠自编码（SAE）网络故障诊断模型。针对SAE网络在特征学习过程出现的梯度硬饱和问题,选用PReLU激活函数,在SAE网络中加入批标准化（BN）层进行优化,通过输出层的Softmax函数,得到变桨系统各部件故障发生概率。以均方误差最小化为目标,采用Adam算法迭代训练数据,使模型参数得到更新。在风电机组变桨系统数据采集与监视控制（SCADA）系统中的数据集中,对优化前后的SAE网络通过改变迭代次数、样本数量进行实验,结果表明,优化后的SAE网络模型具有更好的识别精度;另外,在不同样本数量的实验中,与其他传统模型相比,优化后的SAE网络模型故障识别率也更高,表明其在风电机组故障诊断领域有一定的应用价值。
Abstract
In order to improve the accuracy of fault diagnosis of wind turbine pitch system, a fault diagnosis model based on batch normalization of stacked auto-encode (SAE) network is proposed. Aiming at the problem of hard gradient saturation in the feature learning process of the SAE network, the PReLU activation function is selected, and the batch normalization (BN) layer is added to the SAE network for optimization. Through the Softmax function of the output layer, the failure probability of each component of the pitch system is obtained. With the goal of minimizing the mean square error, the Adam algorithm is used to iterate the training data to update the model parameters. In the data set of wind turbine pitch system supervisory control and data acquisition (SCADA) system, the SAE network before and after optimization is tested by changing the number of iterations and the number of samples. The results show that the optimized SAE network model has better recognition accuracy. In addition, in the experiments with different sample numbers, compared with other traditional models, the fault recognition rate of the optimized SAE network model is also higher, indicating that it has certain application value in the field of wind turbine fault diagnosis.
关键词
风电机组 /
变桨系统 /
故障诊断 /
批标准化 /
堆叠自编码
Key words
wind turbines /
pitch system /
fault diagnosis /
batch normalization /
stacking auto-encoder
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 尹诗, 余忠源, 孟凯峰, 等. 基于非线性状态估计的风电机组变桨控制系统故障识别[J]. 中国电机工程学报, 2014, 34(增刊): 160-165.
YIN S, YU Z Y, MENG K F, et al.Fault identification of pitch control system of wind turbine based on nonlinear state estimation[J]. Proceedings of the CSEE, 2014, 34(S): 160-165.
[2] QIU Y N, FENG Y H, SUN J, et al.Applying thermophysics for wind turbine drivetrain fault diagnosis using SCADA data[J]. IET renewable power genernation, 2016, 10(5): 661-668.
[3] DAI J C, YANG X, HU W, et al.Effect investigation of yaw on wind turbine performance based on SCADA data[J]. Energy, 2018, 149: 684-696.
[4] CHENG F Z, PENG Y Y, QU L Y, et al.Current-based fault detection and identification for wind turbine drivetrain gearboxes[J]. IEEE transactions on industry applications , 2017, 53(2): 878-887.
[5] VÁSQUEZ S, KINNAERT M, PINTELON R. Active fault diagnosis on hydraulic pitch system based on frequency-domain identification[J]. IEEE transactions on control systems technology, 2019, 27(2): 663-678.
[6] 李辉, 杨超, 李学伟, 等. 风机电动变桨系统状态特征参量挖掘及异常识别[J]. 中国电机工程学报, 2014, 34(12): 1922-1930.
LI H, YANG C, LI X W, et al.Conditions characteristic parameters mining and outlier identification for electric pitch system of wind turbine[J]. Proceedings of the CSEE, 2014, 34(12): 1922-1930.
[7] 赵洪山, 闫西慧, 王桂兰, 等. 应用深度自编码网络和XGBoost的风电机组发电机故障诊断[J]. 电力系统自动化, 2019, 43(1): 81-90.
ZHAO H S, YAN X H, WANG G L, et al.Fault diagnosis of wind turbine generator based on deep autoencoder network and XGBoost[J]. Automation of electric power systems, 2019, 43(1): 81-90.
[8] MITRA S, KOLEY C.An automated SCADA based system for identification of induction motor bearing fault used in process control operation[C]//2nd International Conference on Control, Instrumentation, Energy & Communication (CIEC), Kolkata, India, 2016: 294-298.
[9] 丛伟, 胡亮亮, 孙世军, 等. 基于改进深度降噪自编码网络的电网气象防灾方法[J]. 电力系统自动化, 2019, 43(2): 42-50.
CONG W, HU L L, SUN S J, et al.Meteorological disaster prevention method for power grid based on improved stacked denoising autoen-coder network[J].Automation of electric power systems, 2019, 43(2):42-50.
[10] 刘辉海, 赵星宇, 赵洪山, 等. 基于深度自编码网络模型的风电机组齿轮箱故障检测[J]. 电工技术学报, 2017, 32(17): 156-163.
LIU H H, ZHAO X Y, ZHAO H S, et al.Fault detection of wind turbine gearbox based on deep auto-encoder network[J]. Transactions of China Electrotechnical Society, 2017, 32(17): 156-163.
[11] 张西宁, 向宙, 夏心锐, 等. 堆叠自编码网络性能优化及其在滚动轴承故障诊断中的应用[J]. 西安交通大学报, 2018, 52(10): 49-56, 87.
ZHANG X N, XIANG Z, XIA X R, et al.Optimization of stacking auto-encoder with applications in bearing fault diagnosis[J]. Journal of Xi'an Jiaotong University, 2018, 52(10): 49-56, 87.
[12] RUMELHART D E, HINTON G E, WILLIAMS R J.Learning representation by back-propagating errors[J]. Nature, 1988, 323(6088): 399-421.
[13] HINTON G E, SALAKHUTDINOV R R.Reducing the dimensionality of data with neural networ[J]. Science, 2006, 313(5786): 504-507.
[14] 赵洪山, 刘辉海, 刘宏杨, 等. 基于堆叠自编码网络的风电机组发电机状态监测与故障诊断[J]. 电力系统自动化, 2018, 42(11): 102-108.
ZHAO H S, LIU H H, LIU H Y, et al.Condition monitoring and fault diagnosis of wind turbine generator based on stacked auto-encoder network[J]. Automation of electric power systems, 2018, 42(11): 102-108.
[15] WANG L, ZHANG Z J, XU J, et al.Wind turbine blade breakage monitoring with deep autoencoders[J]. IEEE transactions on smart grid, 2018, 9(4): 2824-2833.
[16] 赵志勇. Python机器学习算法[M]. 北京: 电子工业出版社, 2017.
ZHAO Z Y.Python machine learning algorithm[M].Beijing: Electronic Idustry Press, 2017.
[17] 杨云, 杜飞. 深度学习实战[M]. 北京, 清华大学出版社, 2018.
YANG Y, DU F.Deep learning in action[M]. Beijing: Tsinghua University Press, 2018.
[18] KINGMA D P, BA J L.Adam: a method for stochastic optimization[C]//Proceedings of the 3rd International Conference on Learning Representation, 2015.
[19] IOFFE S, SZEGEDY C.Batch normalization: accelerating deep network training by reducing internal covariate shift[C]//32nd International Conference for Learning Representations, San Diego, 2015.
[20] 郭婷婷. 风电机组变工况变桨系统异常状态在线识别[D]. 北京: 北京交通大学, 2019.
GUO T T.On-line abnormal state identification of pitch system based on transitional mode for wind turbine[D]. Beijing: Beijing Jiaotong University, 2019.
基金
国家自然科学基金（51767014; 51867013）
PDF(1703 KB)
