PCA-CLUSTER和EMD-CNN相结合的光伏发电设备故障诊断方法FAULT DIAGNOSIS METHOD OF PV POWER GENERATION EQUIPMENT COMBINED WITH PCA-CLUSTER AND EMD-CNN
裴刘生,周双全,王海峰,赵华鸿
摘要(Abstract):
针对光伏发电设备的输出特性具有时间序列的特征,提出了一种主成分分析-聚类算法(PCACLUSTER)和经验模态分解-卷积神经网络(EMD-CNN)相结合的光伏发电设备故障诊断方法。首先,通过对时间序列进行主成分分析(PCA),从冗余特征中提取主要成分,降低聚类输入维数,再利用K-Means算法对时间序列进行聚类;然后,通过经验模态分解(EMD)方法提取时间序列特征,再利用卷积神经网络(CNN)进行训练和分类,并最终判断出光伏发电设备具体的故障类型。实验结果和应用效果表明,该方法可以有效实现光伏发电设备的故障诊断。
关键词(KeyWords): 光伏发电设备;主成分分析;K-Means算法;经验模态分解;卷积神经网络;故障诊断
基金项目(Foundation):
作者(Author): 裴刘生,周双全,王海峰,赵华鸿
DOI: 10.19911/j.1003-0417.tyn20200601.02
参考文献(References):
- [1]陆灵,朱红路,连魏魏,等.模糊C均值聚类在光伏阵列故障样本数据识别中的应用[J].发电与空调, 2018,39(2):188-194.
- [2]周大镯.多变量时间序列研究[M].石家庄:河北人民出版社, 2006.
- [3]程泽,李兵峰,刘力,等.一种新型结构的光伏阵列故障检测方法[J].电子测量与仪器学报, 2010, 24(2):131-136.
- [4]王元章,李智华,吴春华.光伏系统故障诊断方法综述[J].电源技术, 2013, 37(9):1700-1705.
- [5]梅长林,周家良.实用统计方法[M].北京:科学出版社,2002.
- [6] MARQUES J P. Pattern recognition——Concepts, methods and applications(2nd ed.)[M]. WU Y F, translate. Beijing:Tsinghua University Press, 2002.
- [7] FAHIM A M, SALEM A M, TORKER F A, et al. An efficient enhanced k-means clustering algorithm[J]. Journal of Zhejiang university science A(Science in engineering),2006, 7(10):1626-1633.
- [8] HUANG N E, SHEN Z, LONG S R, et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis[J]. Proceedings:Mathematical, physical and engineering sciences, 1998(1):903-995.
- [9] SRIVASTAVA N, HINTON G, KRIZHEVSKY A, et al.Dropout:A simple way to prevent neural networks from overfitting[J]. Journal of machine learning research,2014(15):1929-1958.