针对风电叶片在复杂工况下面临的跨域数据分布偏移导致的异常状态评测泛化能力不足问题,提出一种动态域自适应残差注意力网络（DDA-RAN）的跨域异常状态评测方法。该方法融合迁移学习与动态注意力机制,通过双向动态校准策略实现领域偏移协同优化。具体而言,设计动态域自适应注意力残差块（DDAAR-Block）,利用动态域自适应注意力动态调整机制结合可切换式域标识,自适应激活源域或目标域分支,平衡领域共性与特异性特征表征;构建跨域对齐模块,通过梯度截断与门控机制实现分阶段动态训练,促进跨域特征交互与分布对齐;采用自适应多尺度残差块提取深层与浅层时序特征,结合全局自适应池化增强特征鲁棒性;进一步引入正则化多层感知器（MLP）分类器,提升模型泛化能力。实验结果表明,在叶片开裂与覆冰跨域评测任务中,所提方法在目标域准确率（最高达99.36%）、F1分数（99.35%）及马修斯相关系数（MCC）（98.71%）等指标上均显著优于主流对比模型（如对抗判别域适应（ADDA）、域对抗神经网络（DANN））,其中准确率平均提升2.1%~3.8%,可为工业场景下的跨域异常评测提供有益参考。
Abstract
To address the insufficient generalization ability of abnormal state assessment caused by cross-domain data distribution shifts in wind turbine blades under complex operating conditions, this paper proposes a Dynamic Domain Adaptive Residual Attention Network （DDA-RAN） based cross-domain anomaly assessment method. This approach integrates transfer learning with a dynamic attention mechanism, achieving collaborative optimization of domain shift through a bidirectional dynamic calibration strategy. Specifically, a Dynamic Domain Adaptive Attention Residual Block （DDAAR-Block） is designed. It utilizes a dynamic domain adaptive attention mechanism combined with switchable domain identifiers to adaptively activate either the source or target domain branch, balancing the representation of domain-invariant and domain-specific features. A cross-domain alignment module is constructed, implementing phased dynamic training via gradient truncation and a gating mechanism to promote cross-domain feature interaction and distribution alignment. An adaptive multi-scale residual block is adopted to extract deep and shallow temporal features, enhanced by global adaptive pooling for feature robustness. Furthermore, a regularized Multi-Layer Perceptron （MLP） classifier is introduced to improve model generalization. Experimental results demonstrate that in cross-domain assessment tasks for blade cracking and icing, the proposed method significantly outperforms mainstream comparison models （e.g., ADDA, DANN） in terms of target domain accuracy （up to 99.36%）, F1-score （99.35%）, and MCC （98.71%）. The accuracy shows an average improvement of 2.1%-3.8%. This provides a valuable reference for cross-domain anomaly assessment in industrial scenarios.
关键词
风电机组叶片 /
迁移学习 /
残差神经网络 /
状态评测 /
动态域自适应 /
跨域对齐
Key words
wind turbine blades /
transfer learning /
residual neural networks /
state assessment /
dynamic domain adaptation /
cross-domain alignment
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 陈雪峰, 郭艳婕, 许才彬, 等. 风电装备故障诊断与健康监测研究综述[J]. 中国机械工程, 2020, 31(2): 175-189.
Chen X F, Guo Y J, Xu C B, et al.Review of fault diagnosis and health monitoring for wind power equipment[J]. China Mechanical Engineering, 2020, 31(2): 175-189.
[2] 安宗文, 马军霞, 马强, 等. 风电叶片全尺寸疲劳试验加载方法研究[J]. 太阳能学报, 2021, 42(10): 250-257.
An Z W, Ma J X, Ma Q, et al.Study on loading scheme for full-scale fatigue test of wind turbine blades[J]. Acta Energiae Solaris Sinica, 2021, 42(10): 250-257.
[3] JIN X H, LYU S Y, KONG Z Q, et al.Graph spatio-temporal networks for condition monitoring of wind turbine[J]. IEEE Transactions on Sustainable Energy, 2024, 15(4): 2276-2286.
[4] 王中行, 周元贵, 张学广. 基于人工智能算法的风电机组状态监测和故障诊断技术研究综述[J]. 东北电力大学学报, 2024, 44(1): 42-51.
Wang Z X, Zhou Y G, Zhang X G.Review of artificial intelligence algorithms-based wind turbine condition monitoring and fault diagnosis techniques[J]. Journal of Northeast Electric Power University, 2024, 44(1): 42-51.
[5] 乔卉卉, 赵二贤, 郝如江, 等. 基于多尺度CNN与双阶段注意力机制的轴承工况域泛化故障诊断[J]. 振动与冲击, 2025, 44(2): 267-278.
Qiao H H, Zhao E X, Hao R J, et al.Operating condition domain generalization fault diagnosis method for bearings based on a multiscale convolutional neural network and the two-stage attention mechanism[J]. Journal of Vibration and Shock, 2025, 44(2): 267-278.
[6] 黄荣舟, 汤宝平, 杨燕妮, 等. 基于长短时记忆网络融合SCADA数据的风电齿轮箱状态监测[J]. 太阳能学报, 2021, 42(1): 235-239.
Huang R Z, Tang B P, Yang Y N, et al.Condition monitoring of wind turbine gearbox based on LSTM neural network fusing SCADA data[J]. Acta Energiae Solaris Sinica, 2021, 42(1): 235-239.
[7] 李康, 李爽, 高小永, 等. 多变量时序标记Transformer及其在电潜泵故障诊断中的应用[J]. 控制与决策, 2025, 40(4): 1145-1153.
Li K, Li S, Gao X Y, et al.Multivariate time series tokenized Transformer and its application in fault diagnosis of electric submersible pump[J]. Control and Decision, 2025, 40(4): 1145-1153.
[8] 邓林峰, 王琦, 郑玉巧. 基于改进残差网络的风电轴承故障迁移诊断方法[J]. 振动工程学报, 2024, 37(2): 356-364.
Deng L F, Wang Q, Zheng Y Q.A fault transfer diagnosis method for wind turbine bearings based on improved residual neural networks[J]. Journal of Vibration Engineering, 2024, 37(2): 356-364.
[9] 李东泽, 齐咏生, 刘利强, 等. 基于双流生成对抗网络数据增强的风电机组智能故障诊断[J]. 电力自动化设备, 2024, 44(11): 94-102.
Li D Z, Qi Y S, Liu H Q, et al.Intelligent fault diagnosis of wind turbine unit based on dual-stream generative adversarial network data augmentation[J]. Electric Power Automation Equipment, 2024, 44(11): 94-102.
[10] 黄星华, 吴天舒, 杨龙玉,等. 一种面向旋转机械的基于Transformer特征提取的域自适应故障诊断[J]. 仪器仪表学报, 2022, 43(11): 210-218.
Huang X H, Wu T S, Yang L Y, et al.Domain adaptive fault diagnosis based on Transformer feature extraction for rotating machinery[J]. Chinese Journal of Scientific Instrument, 2022, 43(11): 210-218.
[11] Yang J Y, An W Z, Yan C C, et al.Context-aware domain adaptation in semantic segmentation[C]//2021 IEEE Winter Conference on Applications of Computer Vision (WACV). Waikoloa, HI, USA, 2021: 514-524.
[12] 贾峰, 李世豪, 沈建军, 等. 采用深度迁移学习与自适应加权的滚动轴承故障诊断[J]. 西安交通大学学报, 2022, 56(8): 1-10.
Jia F, Li S H, Shen J J, et al.Fault diagnosis of rolling bearings using deep transfer learning and adaptive weighting[J]. Journal of Xi’an Jiaotong University, 2022, 56(8): 1-10.
[13] 安文杰, 陈长征, 田淼, 等. 基于迁移学习的风电机组轴承故障诊断研究[J]. 太阳能学报, 2023, 44(6): 367-373.
An W J, Chen C Z, Tian M, et al.Research on bearing fault diagnosis of wind turbines based on transfer learning[J]. Acta Energiae Solaris Sinica, 2023, 44(6): 367-373.
[14] 吕游, 封烁, 郑茜, 等. 基于LeNet5like的迁移学习风电机组叶片覆冰故障诊断研究[J]. 仪器仪表学报, 2024, 45(3): 128-143.
Lyu Y, Feng S, Zheng Q, et al.Research on fault diagnosis of wind turbine icing characteristics based on LeNet5like transfer learning[J]. Chinese Journal of Scientific Instrument, 2024, 45(3): 128-143.
[15] LIU F X, LI Z J, LIANG M L, et al.Prediction method of non-stationary random vibration fatigue reliability of turbine runner blade based on transfer learning[J]. Reliability Engineering & System Safety, 2023, 235: 109215.
[16] Yue R X,Jiang G Q, Jin X H, et al.Spatio-temporal feature alignment transfer learning for cross-turbine blade icing detection of wind turbines[J]. IEEE Transactions on Instrumentation and Measurement, 2024, 73:1-17.
[17] Yang W Q, Yu G.Federated multi-model transfer learning-based fault diagnosis with peer-to-peer network for wind turbine cluster[J]. Machines, 2022, 10(11): 972.
基金
国家自然科学基金（62161042）; 内蒙古自治区自然科学基金（2024LHMS06002）; 高校基本科研业务费（2024QNJ005）
PDF(2846 KB)
