采用变分网络的风力发电机组叶片结冰程度量化方法

王倩月; 司刚全; 尹军华; 童文瀚; 李博; 王昕

doi:10.7652/xjtuxb202508019

您当前的位置：

首页 >

文章列表页 >

采用变分网络的风力发电机组叶片结冰程度量化方法

—————————————————————— | 更新时间：2025-07-30

- 采用变分网络的风力发电机组叶片结冰程度量化方法
- Quantification Method for Ice Accretion on Wind Turbine Blades Using Variational Network
- 西安交通大学学报 2025年59卷第8期页码：199-210
- 作者机构：
  
  1.西安交通大学电工材料电气绝缘全国重点实验室， 710049，西安
  2.西安交通大学电气工程学院， 710049，西安
- 作者简介：
  
  王倩月（1999—），女，博士生
  司刚全，男，教授，博士生导师。
- 基金信息：
  
  国家重点研发计划资助项目(2021YFB2400800;2018AAA0101501)
- DOI：10.7652/xjtuxb202508019
  中图分类号： TP29
- 收稿：2024-12-30，
  
  网络首发：2025-04-10，
  
  纸质出版：2025-08-10
- 稿件说明：
移动端阅览
王倩月, 司刚全, 尹军华, 等. 采用变分网络的风力发电机组叶片结冰程度量化方法[J]. 西安交通大学学报, 2025,59(8):199-210.

WANG Qianyue, SI Gangquan, YIN Junhua, et al. Quantification Method for Ice Accretion on Wind Turbine Blades Using Variational Network[J]. Journal of Xi’an Jiaotong University, 2025, 59(8): 199-210.
王倩月, 司刚全, 尹军华, 等. 采用变分网络的风力发电机组叶片结冰程度量化方法[J]. 西安交通大学学报, 2025,59(8):199-210. DOI： 10.7652/xjtuxb202508019.

WANG Qianyue, SI Gangquan, YIN Junhua, et al. Quantification Method for Ice Accretion on Wind Turbine Blades Using Variational Network[J]. Journal of Xi’an Jiaotong University, 2025, 59(8): 199-210. DOI： 10.7652/xjtuxb202508019.

摘要

针对现有变分网络在风力发电机组监测数据特征提取和故障敏感性方面的不足，提出了一种面向风机正常状态建模的条件变分递归窗口网络（CVRWN），并据此构建叶片结冰程度量化方法。首先，采用引入自注意力机制的多组条件变分自编码器（CVAE）对窗口化的风机监测数据进行学习，以实现各时间窗口的潜在分布特征提取与风功率数据重构；其次，通过门控循环单元（GRU）实现跨窗口特征传递，并引入预测子模块以增强长期趋势建模能力，从而构建完整的CVRWN网络结构；随后，利用风机正常运行数据对所构建网络进行训练，通过联合优化重构损失与预测损失得到具备稳定建模能力的CVRWN模型；最终，将CVRWN模型最后一个窗口的风功率重构损失定义为风机叶片结冰指数

，实现叶片结冰程度的精准量化。实际运行数据验证表明，采用CVRWN模型进行风功率重构时，其精度较基础CVAE模型提升约10%，验证了所提模型结构改进的合理性与有效性。在叶片结冰过程中，相较于基线模型，所提模型的结冰指数

能够精准动态表征叶片结冰演化过程，为极端环境下风力发电机组的安全运维提供量化参考。

Abstract

To address the limitation of existing variational networks in feature extraction and fault sensitivity for wind turbine monitoring data

this paper proposes a conditional variational recurrent window network (CVRWN) aimed at modeling the normal operating state of wind turbines

and constructs a quantification method for ice accretion on blades based on such network. First

multiple groups of conditional variational autoencoders (CVAEs) enhanced with self-attention mechanisms are employed to learn from windowed monitoring data

enabling the extraction of latent distribution features and the reconstruction of wind power data for each time window. Next

gated recurrent units (GRUs) are used to facilitate cross-window feature transfer

and a predictive submodule is introduced to enhance long-term trend modeling capability

thereby constructing the complete CVRWN architecture. Subsequently

the constructed network is trained on turbine monitoring data under normal operating conditions

resulting in a stable CVRWN model obtained by jointly optimizing reconstruction and prediction losses. Finally

the reconstruction loss of wind power in the last window of the CVRWN model is defined as the ice accretion indicator

which enables accurate quantification of the icing levels. Validation with actual operational data demonstrates that the CVRWN model improves wind power reconstruction accuracy by approximately 10% compared to the basic CVAE model

verifying the rationality and effectiveness of the proposed structural enhancements. During the ice accretion process

the proposed model's ice accretion indicator

can accurately and dynamically represent the evolution of blade icing compared to baseline models

providing a quantitative reference for the sa

fe operation and maintenance of wind turbines in extreme environments.

关键词

Keywords

references

龙寰 , 杨婷 , 徐劭辉 , 等 . 基于数据驱动的风电机组状态监测与故障诊断技术综述 [J ] . 电力系统自动化 , 2023 , 47 ( 23 ): 55 - 69 .

LONG Huan , YANG Ting , XU Shaohui , et al . Review of data-driven condition monitoring and fault diagnosis technologies for wind turbines [J ] . Automation of Electric Power Systems , 2023 , 47 ( 23 ): 55 - 69 .

胡琴 , 王欢 , 舒立春 , 等 . 覆冰条件下风力发电机叶片防/除冰方法综述 [J ] . 电工技术学报 , 2024 , 39 ( 17 ): 5482 - 5496 .

HU Qin , WANG Huan , SHU Lichun , et al . Review of anti-/de-icing methods for wind turbine blades under icing conditions [J ] . Transactions of China Electrotechnical Society , 2024 , 39 ( 17 ): 5482 - 5496 .

YANG Li , LI Gaoyang , ZHANG Zihan , et al . Operations & maintenance optimization of wind turbines integrating wind and aging information [J ] . IEEE Transactions on Sustainable Energy , 2021 , 12 ( 1 ): 211 - 221 .

BADIHI H , ZHANG Youmin , JIANG Bin , et al . A comprehensive review on signal-based and model-based condition monitoring of wind turbines: fault diagnosis and lifetime prognosis [J ] . Proceedings of the IEEE , 2022 , 110 ( 6 ): 754 - 806 .

CHENG Xu , SHI Fan , LIU Xiufeng , et al . A novel deep class-imbalanced semisupervised model for wind turbine blade icing detection [J ] . IEEE Transactions on Neural Networks and Learning Systems , 2022 , 33 ( 6 ): 2558 - 2570 .

李特 , 王荣喜 , 高建民 . 风电机组数据采集与监控系统异常数据识别方法 [J ] . 西安交通大学学报 , 2024 , 58 ( 3 ): 106 - 116 .

LI Te , WANG Rongxi , GAO Jianmin . A method for abnormal data recognition of wind turbine supervisory control and data acquisition systems [J ] . Journal of Xi'an Jiaotong University , 2024 , 58 ( 3 ): 106 - 116 .

曾祥军 , 冯琛 , 杨明 , 等 . 考虑运行状态相似性的风电机组数据异常检测方法 [J ] . 电力系统自动化 , 2022 , 46 ( 11 ): 170 - 180 .

ZENG Xiangjun , FENG Chen , YANG Ming , et al . Data anomaly detection method for wind turbines considering operation state similarity [J ] . Automation of Electric Power Systems , 2022 , 46 ( 11 ): 170 - 180 .

王爽心 , 郭婷婷 , 李蒙 . 风电机组变工况变桨系统异常状态在线识别 [J ] . 中国电机工程学报 , 2019 , 39 ( 17 ): 5144 - 5152 .

WANG Shuangxin , GUO Tingting , LI Meng . On-line abnormal state identification of pitch system based on transitional mode for wind turbine [J ] . Proceedings of the CSEE , 2019 , 39 ( 17 ): 5144 - 5152 .

XIANG Ling , WANG Penghe , YANG Xin , et al . Fault detection of wind turbine based on SCADA data analysis using CNN and LSTM with attention mechanism [J ] . Measurement , 2021 , 175 : 109094 .

ZHENG Zhong , ZHANG Zijun . A stochastic recurrent encoder decoder network for multistep probabilistic wind power predictions [J ] . IEEE Transactions on Neural Networks and Learning Systems , 2024 , 35 ( 7 ): 9565 - 9578 .

兰孝升 , 李云凤 , 苏元浩 , 等 . 基于关联度与自检验长短期记忆网络的风电机组轴承寿命预测模型 [J ] . 高电压技术 , 2023 , 49 ( 6 ): 2652 - 2661 .

LAN Xiaosheng , LI Yunfeng , SU Yuanhao , et al . Wind turbine bearing life prediction model based on indexed relation and self-checking long short-term memory [J ] . High Voltage Engineering , 2023 , 49 ( 6 ): 2652 - 2661 .

VASWANI A , SHAZEER N , PARMAR N , et al . Attention is all you need [C ] // Proceedings of the 31st International Conference on Neural Information Processing Systems . Red Hook, NY, USA : Curran Associates Inc. ] , 2017 : 6000 - 6010 .

ENCALADA-DÁVILA Á , MOYÓN L , TUTIVÉN C , et al . Early fault detection in the main bearing of wind turbines based on gated recurrent unit (GRU) neural networks and SCADA data [J ] . IEEE/ASME Transactions on Mechatronics , 2022 , 27 ( 6 ): 5583 - 5593 .

XIAO Jin , LI Chunyan , LIU Bo , et al . Prediction of wind turbine blade icing fault based on selective deep ensemble model [J ] . Knowledge-Based Systems , 2022 , 242 : 108290 .

CHENG Xu , SHI Fan , ZHAO Meng , et al . Temporal attention convolutional neural network for estimation of icing probability on wind turbine blades [J ] . IEEE Transactions on Industrial Electronics , 2022 , 69 ( 6 ): 6371 - 6380 .

翟正利 , 梁振明 , 周炜 , 等 . 变分自编码器模型综述 [J ] . 计算机工程与应用 , 2019 , 55 ( 3 ): 1 - 9 .

ZHAI Zhengli , LIANG Zhenming , ZHOU Wei , et al . Research overview of variational auto-encoders models [J ] . Computer Engineering and Applications , 2019 , 55 ( 3 ): 1 - 9 .

QIN Yi , ZHOU Jianghong , CHEN Dingliang . Unsupervised health indicator construction by a novel degradation-trend-constrained variational autoencoder and its applications [J ] . IEEE/ASME Transactions on Mechatronics , 2022 , 27 ( 3 ): 1447 - 1456 .

张聪 , 朱永生 , 杨敏燕 , 等 . 采用多变量耦合网络与变分图自编码器的机械设备异常检测方法 [J ] . 西安交通大学学报 , 2021 , 55 ( 4 ): 20 - 28 .

ZHANG Cong , ZHU Yongsheng , YANG Minyan , et al . Anomaly detection method with multivariable coupling network and variational graph autoencoder [J ] . Journal of Xi'an Jiaotong University , 2021 , 55 ( 4 ): 20 - 28 .

段雪源 , 付钰 , 王坤 . 基于VAE-WGAN的多维时间序列异常检测方法 [J ] . 通信学报 , 2022 , 43 ( 3 ): 1 - 13 .

DUAN Xueyuan , FU Yu , WANG Kun . Multi-dimensional time series anomaly detection method based on VAE-WGAN [J ] . Journal on Communications , 2022 , 43 ( 3 ): 1 - 13 .

LIN Shuyu , CLARK R , BIRKE R , et al . Anomaly detection for time series using VAE-LSTM hybrid model [C ] // ICASSP 2020-2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . Piscataway, NJ, USA : IEEE , 2020 : 4322 - 4326 .

甘雨 , 郭鹏 , 林立栋 . 基于变分贝叶斯推断的DPGMM风电机组异常数据识别研究 [J ] . 动力工程学报 , 2023 , 43 ( 7 ): 885 - 892 .

GAN Yu , GUO Peng , LIN Lidong . Abnormal wind turbine data identification using a Dirichlet process Gaussian mixture model based on variational Bayesian inference [J ] . Journal of Chinese Society of Power Engineering , 2023 , 43 ( 7 ): 885 - 892 .

吴岚 , 柳亦兵 , 吴仕明 , 等 . 基于VMD倒频变换的风电机组齿轮箱复合故障诊断 [J ] . 振动与冲击 , 2023 , 42 ( 24 ): 221 - 227 .

WU Lan , LIU Yibing , WU Shiming , et al . Composite fault diagnosis of wind turbine gearboxes based on VMD cepstral transform [J ] . Journal of Vibration and Shock , 2023 , 42 ( 24 ): 221 - 227 .

任建亭 , 汤宝平 , 雍彬 , 等 . 基于深度变分自编码网络融合SCADA数据的风电齿轮箱故障预警 [J ] . 太阳能学报 , 2021 , 42 ( 4 ): 403 - 408 .

REN Jianting , TANG Baoping , YONG Bin , et al . Wind turbine gearbox fault warning based on depth variational autoencoders network fusion SCADA data [J ] . Acta Energiae Solaris Sinica , 2021 , 42 ( 4 ): 403 - 408 .

WANG Lei , HE Yigang , SHAO Kaixuan , et al . An unsupervised approach to wind turbine blade icing detection based on beta variational graph attention autoencoder [J ] . IEEE Transactions on Instrumentation and Measurement , 2024 , 73 : 1 - 12 .

POL A A , BERGER V , GERMAIN C , et al . Anomaly detection with conditional variational autoencoders [C ] // 2019 18th IEEE International Conference on Machine Learning and Applications (ICMLA) . Piscataway, NJ, USA : IEEE , 2019 : 1651 - 1657 .

CHUNG J , KASTNER K , DINH L , et al . A recurrent latent variable model for sequential data [C ] // Proceedings of the 29th International Conference on Neural Information Processing Systems . Cambridge, MA, USA : MIT Press , 2015 : 2980 - 2988 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

混合电解槽阵列协同的风光耦合制氢系统双层多目标优化

张角凹坑气膜孔对双层壁气膜冷却特性的影响

隧道内燃料电池汽车氢气泄漏扩散特性及影响因素

扇贝仿生密封腔室内超临界二氧化碳泄漏流湍动能耗散机理研究

超临界二氧化碳向心透平轮背间隙内风阻损失及流动特性研究