基于核主成分分析与长短时记忆网络的水电机组监测预警

PDF(1915 KB)

振动与冲击 ›› 2024, Vol. 43 ›› Issue (24) : 287-294.

论文

基于核主成分分析与长短时记忆网络的水电机组监测预警

王勇飞1，李晓飞1，孙雨欣2，张健1，郭鹏程2，王仁本1

作者信息 +

A monitoring and warning method for hydroelectric units based on KPCA and LSTM

WANG Yongfei1, LI Xiaofei1, SUN Yuxin2, ZHANG Jian1, GUO Pengcheng2, WANG Renben1

Author information +

文章历史 +

摘要

水电机组的可靠稳定运行对于区域电力系统安全极为重要，本文提出了一种基于核主成分分析（Kernel Principal Component Analysis，KPCA）和长短时记忆网络（Long Short-Term Memory，LSTM）的水电机组智能预警方法。首先开展水电机组多通道振动信号数据融合研究，通过KPCA方法去除了多通道信号间冗余，实现了原始数据的压缩表征，并获得了机组在稳态运行工况的T2和SPE控制限，将其作为预警阈值对融合后信号进行异常状态识别。以LSTM为基础构建了时序预测模型，结合异常状态识别结果实现了水电机组状态预警功能。研究通过案例实施验证了所提出方法的有效性，并与KPCA-RNN和KPCA-Informer等模型进行了对比，所提出KPCA-LSTM模型预测结果的R2系数大于0.97，预测偏差处于极低水平，性能优于对比模型。

Abstract

The safe and stable operation of hydropower units is very important for the safety of power stations and regional power grids. An intelligent early warning method for hydropower units based on Kernel Principal Component Analysis (KPCA) and Long Short-Term Memory (LSTM) is proposed in this paper, which realizes that trend early war function based on multi-channel vibration signal fusion. Firstly, the multi-channel vibration signals of hydropower units are fused, and KPCA is used to compress the data to reduce the amount of information processing in the process of early warning. Furthermore, the early warning threshold is set according to the steady-state operation condition of the unit by the multivariate statistical process control method, and the early warning function based on data fusion is realized. Then, the prediction model of vibration fusion data is constructed based on LSTM, and the prediction function of future operation data is realized. In that end, the effectiveness of the propose method is verified by the given case with a R2 coefficient exceeding 0.97, and realize the state monitoring and alarm prediction. Compared to the KPCA-RNN and KPCA-Informer models, the proposed model demonstrates the best performance on the same experimental data.

导出引用

王勇飞1, 李晓飞1, 孙雨欣2, 张健1, 郭鹏程2, 王仁本1. 基于核主成分分析与长短时记忆网络的水电机组监测预警[J]. 振动与冲击, 2024, 43(24): 287-294

WANG Yongfei1, LI Xiaofei1, SUN Yuxin2, ZHANG Jian1, GUO Pengcheng2, WANG Renben1. A monitoring and warning method for hydroelectric units based on KPCA and LSTM[J]. Journal of Vibration and Shock, 2024, 43(24): 287-294

参考文献

[1] 陈鹏, 吴一凡, 蔡爽, 等. 基于自编码压缩与多尺度特征提取的抽水蓄能机组劣化趋势评估与预测[J]. 水利学报, 2022, 53 (06): 747-56.
CHEN Peng, WU Yifan, CAI Shuang, et al. Degradation trend assessment and prediction of pumped storage unit based on deep auto-encoder compression and multiscale feature extraction[J]. Journal of Hydraulic Engineering, 2022, 53 (06): 747-56.
[2] 秦净净, 职保平, 杨春景. 基于混合TPA的水电机组厂房振动传导研究[J]. 振动与冲击, 2023, 42(4): 48-53+80.
QIN Jingjing, ZHI Baoping, YANG Chunjing. Vibration conduction analysis of a hydropower plant based on hybrid TPA. JOURNAL OF VIBRATION AND SHOCK, 2023, 42(4): 48-53.
[3] 付文龙, 周建中, 李超顺, 等. 基于模糊K近邻支持向量数据描述的水电机组振动故障诊断研究[J]. 中国电机工程学报, 2014, 34(32): 5788-5795.
FU Wenlong, ZHOU Jianzhong, LI Chaoshun, XIAO Han, XIAO Jian, ZHU Wenlong. Vibrant Fault Diagnosis for Hydro-Electric Generating Unit Based on Support Vector Data Description Improved With Fuzzy K Nearest Neighbor. Proceedings of the CSEE. 2014, 34(32): 5788-5795
[4] 胡晓, 肖志怀, 刘东, 等. 基于EEMD-SDCC Ⅰ-HMM的水电机组振动故障识别方法[J]. 振动与冲击, 2022, 41(3): 165-175+230.
HU Xiao, XIAO Zhihuai, LIU Dong, et al. Vibration fault identification method of hydropower unit based on EEMD-SDCCⅠ-HMM.JOURNAL OF Vibration and Shock, 2022, 41(3): 165-175.
[5] 毕扬, 郑波, 张亚武, 等. 基于MIC与BiGRU的水电机组振动趋势预测[J]. 水利学报, 2021, 52(5): 612-621+632.
Bi Yang, Zheng Bo, Zhang Yawu, Zhu Xi, Jiang Yalan, Li Chaoshun. Vibration trend prediction of hydropower units based on MIC and BiGRU[J]. Journal of Water Resources, 2021, 52(5): 612-621+632.
[6] LEI Yaguo, LI Naipeng, GUO Liang, et al. Machinery health prognostics: A systematic review from data acquisition to RUL prediction[J]. Mechanical Systems and Signal Processing, 2018, 104: 799-834.
[7] ZHOU, Kaibo, ZHANG Jianyu, SHAN Yahui, et al. A hybrid multi-objective optimization model for vibration tendency prediction of hydropower generators[J]. Sensors, 2019, 19(9): 2055.
[8] Zhao R, Yan R, Chen Z, et al. Deep learning and its applications to machine health monitoring[J]. Mechanical Systems and Signal Processing, 2019, 115: 213-237.
[9] KOUHI Sajjad, KEYNIA Farshid. A new cascade NN based method to short-term load forecast in deregulated electricity market[J]. Energy Conversion and Management, 2013, 71: 76-83.
[10] Kaur Amanpreet, Pedro Hugo T. C., Coimbra Carlos F. M. Ensemble re-forecasting methods for enhanced power load prediction[J]. Energy Conversion and Management, 2014, 80: 582-90.
[11] 张博, 徐卓飞, 李小周, 等. 基于经验模式滤波与循环神经网络的水锤压力信号预测[J]. 排灌机械工程学报, 2022, 40(11): 1120-1125.
ZHANG Bo, XU Zhuofei, LI Xiaozhou, et al. Prediction for water hammer pressure signal based on empirical mode filter and recurrent neural network[J]. Journal of Drainage and Irrigation Machinery Engineering, 2022, 40(11): 1120-1125.
[12] HOCHREITER S, SCHMIDHUBER J J N C. Long short-term memory [J]. Neural computation, 1997, 9(8): 1735-80.
[13] 陈畅, 李晓磊, 崔维玉. 基于LSTM网络预测的水轮机机组运行状态检测[J]. 山东大学学报(工学版), 2019, 49(03): 39-46.
CHEN Chang, LI Xiaolei, CUI Weiyu. Hydraulic turbine operation status detection based on LSTM network prediction[J]. Journal of Shandong University. Engineering Science, 2019, 49(03): 39-46.
[14] Zolfaghari Mehdi, Golabi Mohammad Reza. Modeling and predicting the electricity production in hydropower using conjunction of wavelet transform, long short-term memory and random forest models[J]. Renewable Energy, 2021, 170: 1367-81.
[15] 傅质馨, 殷贵, 朱俊澎,等. 基于EEMD和LSTM的水电机组劣化度预测方法研究[J]. 太阳能学报, 2022, 43(02): 75-81.
FU Zhixin, YIN Gu, ZHU Junpeng, et al. Research on forecasting method of hydropower unit deterioration based on EEMD and LSTM[J]. Acta Energiae Solaris Sinica, 2022, 43(02): 75-81.
[16] 兰家法, 周玉辉, 高泽良, 等. 基于机器学习的水电机组劣化趋势预测模型[J]. 水力发电学报, 2022, 41(12): 135-44.
LAN Jiafa, ZHOU Yuhui, GAO Zeliang， et al. Model for predicting deterioration trends of hydropower units based on machine learning[J]. Journal of Hydroelectric Engineering, 2022, 41(12): 135-44.
[17] 罗毅, 武博翔. 基于深度学习LSTM-DBN的水轮机振动故障预测方法[J]. 振动,测试与诊断, 2022, 42(6): 1233-1238+1251.
Luo Yi, Wu Boxiang. A deep learning LSTM-DBN based method for hydraulic turbine vibration fault prediction[J]. Vibration. Test and Diagnosis, 2022, 42(6): 1233-1238+1251.
[18] 陈飞,王斌,刘婷等.基于时移多尺度注意熵和随机森林的水电机组故障诊断[J].水利学报, 2022, 53(03):358-368+378.
CHEN F, WANG B, LIU T, et al. Fault diagnosis of hydropower units based on time-shifted multiscale attention entropy and random forest [J]. Journal of Hydraulic Engineering, 2022,53(03):358-368+378.
[19] 徐雄, 林海军, 刘悠勇, 等. 融合PCA与自适应K-Means聚类的水电机组故障检测在线方法[J]. 电子测量与仪器学报, 2022, 36(3): 260-267.
XU Xiong, LIN Haijun, LIU Youyong, et al. Online fault detection method of hydraulic turbine combining PCA and adaptive k-means clustering[J]. Journal of Electronic Measurement and Instrumentation, 2022, 36(3): 260-267.
[20] 易辉, 梅磊, 李丽娟, 等. 基于多分类相关向量机的水电机组振动故障诊断[J]. 中国电机工程学报, 2014, 34(17): 2843-2850.
YI Hui, MEI Lei, LI Lijuan, LIU Yufang, YUAN Yuhao. Vibration Fault Diagnosis for Hydroelectric Generating Units Using the Multi-class Relevance Vector Machine. Proceedings of the CSEE. 2014, 34(17): 2843-2850
[21] 潘罗平, 唐波, 周叶, 等. 水电机组故障诊断技术的应用现状与技术发展[J]. 水电站机电技术, 2010, 33(3): 107-109.
Pan Luoping, Tang Bo, Zhou Ye, et al. The current and development of fault diagnosis technologies for hydropower generating unit[J]. Mechanical & Electrical Technique of Hydropower Station, 2010, 33(3): 107-109.
[22] 黎梓昕, 林海军, 徐雄, 等. 基于KPCA-PSO-SVM的水轮机组故障检测方法[J]. 排灌机械工程学报, 2023, 41(5): 467-474.
LI Zixin, LIN Haijun, XU Xiong, et al. Fault detection method of hydraulic turbine unit based on KPCA-PSO-SVM[J]. Journal of Drainage and Irrigation Machinery Engineering, 2023, 41(5): 467-474.
[23] 李志军, 牛晓旭, 李月恒. 基于KPCA-ARIMA模型的TE过程故障预测[J]. 工业控制计算机, 2019, 32(4): 45-48.
TE Process Fault Prediction Based on KPCA－ARIMA Model[J]. Industrial Control Computer, 2019, 32(4): 45-48.
[24] 聂小辉, 金磊. 核主元分析在航天器飞轮自主故障诊断的应用[J]. 北京航空航天大学学报, 2023, 49(8): 2119-2128.
NIE Xiaohui, JIN Lei. Application of kernel principal component analysis in autonomous fault diagnosis for spacecraft flywheel[J]. Journal of Beijing University of Aeronautics and Astronautics, 2023, 49(8): 2119-2128.
[25] 刘爱民, 杨彦军, 王殿学, 等. 动态KPCA-TS在风电齿轮箱安全运行中的应用研究[J]. 控制工程, 2022, 6(29): 1072-1081.
LIU Aiming, YANG Yanjun, WANG Dianxue, et al. Research on application of dynamic KPCA-TS in safe operation of wind turbine gearboxes[J]. Control Engineering of China, 2022, 6(29): 1072-1081.
[26] 卜玮晶,潘树昌. 混流式水轮机偏负荷工况内流场及压力脉动分析[J]. 排灌机械工程学报, 2023, 41(9): 906-912.
BU Weijing, PAN Shuchang. Analysis of internal flow field and pressure pulsation in Francis turbine under partial loa