针对风电场内机组及变电站电气柜热成像温度监测中镜头畸变导致的温度值映射偏差和预测精度低的问题,提出一种基于混合深度神经网络的风电场电气柜温度误差校正和预测方法。该方法利用高斯卷积核结构对含噪图元进行滤波处理,完成图元去噪。建立多层卷积神经网络温度校正模型（CNNs）,将去噪后的图元作为输入,消除图元数据中畸变点导致的测温误差,可提高温度值与目标测点间映射关系的精确度。基于校正赋值后的温度数据建立XGBoost-LSTM加权融合预测模型,输出关键监测点温度预测值可用于风电场设备故障预警中,基于模型结构差异性特点对所加入权重的训练来提高融合模型预测的准确率。算例结果表明该文所提方法可有效校正图元畸变并大幅提高预测精度。
Abstract
A hybrid deep neural network-based temperature correction and prediction method for wind farm electrical cabinets is proposed to address the problems of mapping bias and low prediction accuracy of temperature values caused by lens distortion in thermal imaging temperature monitoring of electrical cabinets in wind farms. The method uses Gaussian convolutional kernel structure to filter noisy image elements so as to complete image element denoising. A multilayer Convolutional Neural Networks (CNNs) model is developed. The denoised image elements are used as input to eliminate the distortion points in the image elements and improve the accuracy of the mapping between the temperature values and the target measurement points. The XGBoost-LSTM weighted fusion prediction model is built based on the corrected and assigned temperature data, and the output temperature prediction values of key monitoring points can be used in wind farm equipment fault warning, and the weights are trained to improve the prediction accuracy of the fusion model based on the structural variability of the model. The results show that the proposed method can effectively correct the distortion of the graphical elements and significantly improve the prediction accuracy.
关键词
风电场 /
温度预测 /
畸变校正 /
卷积神经网络 /
XGBoost /
LSTM
Key words
wind farm /
temperature prediction /
distortion correction /
convolutional neural network /
XGBoost /
LSTM
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 张羽飞, 孟凡勇, 王永千, 等. 基于改进型LSTM的电力设备温度预测方法研究[J]. 电子测量与仪器学报, 2021, 35(12): 167-173.
ZHANG Y F, MENG F Y, WANG Y Q, et al.Research on temperature prediction method of power equipment based on improved LSTM[J]. Journal of electronic measurement and instrumentation, 2021, 35(12): 167-173.
[2] 杨明岱, 王宪, 赵前程, 等. 永磁同步风电机组发电机定子温度建模方法[J]. 太阳能学报, 2022, 43(5): 319-326.
YANG M D, WANG X, ZHAO Q C, et al.Modeling method of stator temperature of PMSG wind turbine[J]. Acta energiae solaris sinica, 2022, 43(5): 319-326.
[3] LI J, SU J, ZENG X.A solution method for image distortion correction model based on bilinear interpolation[J]. Computer optics, 2019, 43(1): 99-104.
[4] WU Y N, FAN Z, FANG Y C, et al.An effective correction method for AFM image distortion due to hysteresis and thermal drift[J]. IEEE transactions on instrumentation and measurement, 2021, 70: 1-12.
[5] 张志强, 王萍, 于旭东, 等. 高精度红外热成像测温技术研究[J]. 仪器仪表学报, 2020, 41(5): 10-18.
ZHANG Z Q, WANG P, YU X D, et al.Study on high accuracy temperature measurement technology of infrared thermal imager[J]. Chinese journal of scientific instrument, 2020, 41(5): 10-18.
[6] 郭鹏, 王桂松, 胥佳, 等. 基于BootStrap置信度计算的风电机组大部件温度预警方法[J]. 太阳能学报, 2021, 42(10): 168-175.
GUO P, WANG G S, XU J, et al.Temperature fault early warning method of wind turbine large components based on BootStrap confidence calculation[J]. Acta energiae solaris sinica, 2021, 42(10): 168-175.
[7] 魏书荣, 张鑫, 符杨, 等. 基于GRA-LSTM-Stacking模型的海上双馈风力发电机早期故障预警与诊断[J]. 中国电机工程学报, 2021, 41(7): 2373-2383.
WEI S R, ZHANG X, FU Y, et al.Early fault warning and diagnosis of offshore wind DFIG based on GRA-LSTM-stacking model[J]. Proceedings of the CSEE, 2021, 41(7): 2373-2383.
[8] 苏连成, 邢美玲, 张慧. 基于组合预测模型的风电机组关键部位故障检测[J]. 太阳能学报, 2021, 42(10): 220-225.
SU L C, XING M L, ZHANG H.Fault detection of key components of wind turbine based on combination prediction model[J]. Acta energiae solaris sinica, 2021, 42(10): 220-225.
[9] DHIMAN H S, DEB D, MUYEEN S M, et al.Wind turbine gearbox anomaly detection based on adaptive threshold and twin support vector machines[J]. IEEE transactions on energy conversion, 2021, 36(4): 3462-3469.
[10] 张文清, 肖厚元, 王众, 等. 基于双核高斯滤波的PSP图像去噪[J]. 电子测试, 2022(1): 39-42.
ZHANG W Q, XIAO H Y, WANG Z, et al.PSP image denoising with duo-core Gaussian filter[J]. Electronic test, 2022(1): 39-42.
[11] ZHANG K, GENG X P, YAN X H.Prediction of 3-D ocean temperature by multilayer convolutional LSTM[J]. IEEE geoscience and remote sensing letters, 2020, 17(8): 1303-1307.
[12] FU J, CHU J C, GUO P, et al.Condition monitoring of wind turbine gearbox bearing based on deep learning model[J]. IEEE access, 2019, 7: 57078-57087.
[13] 曹力, 潘巧波, 王明宇, 等. 基于混合核函数支持向量机的风电机组发电机温度预警方法[J]. 华电技术, 2020, 42(5): 43-49.
CAO L, PAN Q B, WANG M Y, et al.Early warning method for wind turbine generator temperature based on HK-SVM[J]. Huadian technology, 2020, 42(5): 43-49.
基金
国家自然科学基金（52007124）; 辽宁省揭榜挂帅科技攻关专项（2021JH1/10400009）
PDF(2607 KB)
