基于改进卷积神经网络的风电机组叶片覆冰诊断方法研究

邢作霞; 张玥; 郭珊珊; 张超

doi:10.19912/j.0254-0096.tynxb.2023-1953

PDF(2024 KB)

太阳能学报 ›› 2025, Vol. 46 ›› Issue (3) : 661-667. DOI: 10.19912/j.0254-0096.tynxb.2023-1953

基于改进卷积神经网络的风电机组叶片覆冰诊断方法研究

邢作霞^1,2, 张玥¹, 郭珊珊^1,2, 张超^1,2

作者信息 +

DIAGNOSIS OF BLADE ICING BASED ON IMPROVED CONVOLUTION NEURAL NETWORK IN WIND TURBINE STUDY

Xing Zuoxia^1,2, Zhang Yue¹, Guo Shanshan^1,2, Zhang Chao^1,2

Author information +

文章历史 +

摘要

针对风电机组叶片覆冰影响机组运行安全和降低发电量的问题,提出一种基于极端梯度提升算法和麻雀搜索算法优化卷积神经网络的风电机组叶片覆冰诊断方法。首先,利用基于极端梯度提升算法计算实际机组监控和数据采集系统（SCADA）数据的特征权重,筛除冗余特征变量,降低诊断模型的复杂度、减少诊断时间;再利用卷积神经网络模型对筛选后SCADA数据进行特征提取建立叶片覆冰诊断分类模型;最后,利用麻雀搜索算法对诊断模型中的超参数寻优,提高诊断模型的准确率。实验结果表明提出的方法对叶片覆冰的诊断准确率达到98%,相比于长短期记忆网络、K近邻算法等分类模型诊断准确率更高。

Abstract

Wind turbine blade icing poses significant risks to operational safety and energy production efficiency. To address this challenge, this study proposes an ice detection framework combining XGBoost feature selection with a Sparrow Search Algorithm （SSA）-optimized convolutional neural network （CNN）. First, XGBoost evaluates feature importance in SCADA system data to eliminate redundant variables, streamlining model architecture and accelerating diagnostic processes. A CNN then extracts discriminative features from this refined dataset to establish an ice accretion classification system. The SSA subsequently optimizes critical hyperparameters in the CNN model to enhance detection precision. Experimental validation demonstrates 98% diagnostic accuracy for blade icing conditions, outperforming both Long Short-Term Memory networks（97.2%） and k-Nearest Neighbors classifiers （95.1%）. This integrated approach provides a reliable solution for real-time ice monitoring in wind farms.

导出引用

邢作霞, 张玥, 郭珊珊, 张超. 基于改进卷积神经网络的风电机组叶片覆冰诊断方法研究[J]. 太阳能学报. 2025, 46(3): 661-667 https://doi.org/10.19912/j.0254-0096.tynxb.2023-1953

Xing Zuoxia, Zhang Yue, Guo Shanshan, Zhang Chao. DIAGNOSIS OF BLADE ICING BASED ON IMPROVED CONVOLUTION NEURAL NETWORK IN WIND TURBINE STUDY[J]. Acta Energiae Solaris Sinica. 2025, 46(3): 661-667 https://doi.org/10.19912/j.0254-0096.tynxb.2023-1953

中图分类号： TK81

参考文献

[1] CAI C, GUO J C, SONG X W, et al.Review of data-driven approaches for wind turbine blade icing detection[J]. Sustainability, 2023, 15(2): 1617.
[2] WANG X, ZHENG Z, JIANG G Q, et al.Detecting wind turbine blade icing with a multiscale long short-term memory network[J]. Energies, 2022, 15(8): 2864.
[3] QU B Y, XING Z X, LIU Y, et al.Research on short-term output power forecast model of wind farm based on neural network combination algorithm[J]. Wind energy, 2022, 25(10): 1710-1734.
[4] XING Z X, CHEN M Y, CUI J, et al.Detection of magnitude and position of rotor aerodynamic imbalance of wind turbines using convolutional neural network[J]. Renewable energy, 2022, 197: 1020-1033.
[5] DU Y, ZHOU S X, JING X J, et al.Damage detection techniques for wind turbine blades: a review[J]. Mechanical systems and signal processing, 2020, 141: 106445.
[6] 张莹博, 单光华, 王飞, 等. 风电机组叶片覆冰预测和防覆冰技术综述[J]. 复合材料科学与工程, 2024(8): 119-128.
ZHANG Y B, SHAN G H, WANG F, et al.Review of prediction and prevention techniques for wind turbine blade icing[J]. Composites science and engineering, 2024(8): 119-128.
[7] 高峰, 张鸿, 许琳, 等. 风电机组叶片故障仿真与状态判别研究[J]. 太阳能学报, 2023, 44(4): 52-59.
GAO F, ZHANG H, XU L, et al.Research on fault simulation and state judgment of wind turbine blades[J]. Acta energiae solaris sinica, 2023, 44(4): 52-59.
[8] 朱恩龙, 冯聪聪, 申振腾, 等. 基于机器视觉的风力机叶片故障检测技术研究[J]. 太阳能学报, 2023, 44(4): 209-215.
ZHU E L, FENG C C, SHEN Z T, et al.Research on wind turbine blade fault detection technology based on machine vision[J]. Acta energiae solaris sinica, 2023, 44(4): 209-215.
[9] 王道累, 肖佳威, 刘易腾, 等. 风电机组叶片损伤检测技术研究与进展[J]. 中国电机工程学报, 2023, 43(12): 4614-4631.
WANG D L, XIAO J W, LIU Y T, et al.Research and development of wind turbine blade damage detection technology[J]. Proceedings of the CSEE, 2023, 43(12): 4614-4631.
[10] 刘家瑞, 杨国田, 杨锡运. 基于深度卷积自编码器的风电机组故障预警方法研究[J]. 太阳能学报, 2022, 43(11): 215-223.
LIU J R, YANG G T, YANG X Y.Research on wind turbine fault warning method based on deep convolution auto-encoder[J]. Acta energiae solaris sinica, 2022, 43(11): 215-223.
[11] LIU Y, CHENG H, KONG X G, et al.Intelligent wind turbine blade icing detection using supervisory control and data acquisition data and ensemble deep learning[J]. Energy science & engineering, 2019, 7(6): 2633-2645.
[12] JIANG G Q, LI W Y, BAI J R, et al.SCADA data-driven blade icing detection for wind turbines: an enhanced spatio-temporal feature learning approach[J]. Measurement science and technology, 2023, 34(5): 054004.
[13] TONG R N, LI P, GAO L, et al.A novel ellipsoidal semisupervised extreme learning machine algorithm and its application in wind turbine blade icing fault detection[J]. IEEE transactions on instrumentation and measurement, 2022, 71: 3523116.
[14] 王永平, 张蕾, 张晓琳, 等. 基于CNN-BiGRU的风机叶片故障诊断[J]. 内蒙古科技大学学报, 2022, 41(2): 173-179.
WANG Y P, ZHANG L, ZHANG X L, et al.Fault diagnosis of wind turbine blades based on CNN-BiGRU[J]. Journal of Inner Mongolia university of science and technology, 2022, 41(2): 173-179.
[15] 沈学利, 杨莹, 秦鑫宇, 等. 基于残差神经网络的风机叶片结冰故障诊断[J]. 噪声与振动控制, 2022, 42(1): 79-87.
SHEN X L, YANG Y, QIN X Y, et al.Icing fault diagnosis of wind turbine blades based on residual neural network[J]. Noise and vibration control, 2022, 42(1): 79-87.
[16] 汤占军, 史小兵, 肖遥, 等. 基于深度学习模型的风机叶片结冰故障诊断[J]. 噪声与振动控制, 2023, 43(4): 96-103.
TANG Z J, SHI X B, XIAO Y, et al.Icing fault diagnosis of wind turbine blades based on deep learning model[J]. Noise and vibration control, 2023, 43(4): 96-103.
[17] ANG J C, MIRZAL A, HARON H, et al.Supervised, unsupervised, and semi-supervised feature selection: a review on gene selection[J]. IEEE/ACM transactions on computational biology and bioinformatics, 2016, 13(5): 971-989.
[18] YANG X Y, HUANG X X, GAO X X, et al.Icing diagnosis model for wind turbine blade based on feature optimization and 1D-convolutional neural network[J]. 2022, 14(3): 033303.
[19] 任东方, 马家庆, 何志琴, 等. 基于AVMD-CNN-GRU-Attention的超短期风功率预测研究[J]. 太阳能学报, 2024, 45(6): 436-443.
REN D F, MA J Q, HE Z Q, et al.Research on ultra-short-term wind power forecast based on AVMD-CNN-GRU-Attention[J]. Acta energiae solaris sinica, 2024, 45(6): 436-443.
[20] LI Q, SHI Y R, LIN R Q, et al.A novel oil pipeline leakage detection method based on the sparrow search algorithm and CNN[J]. Measurement, 2022, 204: 112122.