a novel sensor fault diagnosis method based on modified ensemble empirical mode decomposition and probabilistic neural network论文.pdf

Measurement 68 (2015) 328–336 Contents lists available at ScienceDirect Measurement journal homepage: /locate/measurement A novel sensor fault diagnosis method based on Modified Ensemble Empirical Mode Decomposition and Probabilistic Neural Network Yunluo Yu, Wei Li ⇑, Deren Sheng, Jianhong Chen Institute of Thermal Science and Power System, Zhejiang University, Hangzhou 310027, Zhejiang Province, China a r t i c l e i n f o a b s t r a c t Article history: A novel fault diagnosis method based on Modified Ensemble Empirical Mode Decomposition Received 28 May 2014 (MEEMD) and Probabilistic Neural Network (PNN) is presented in this paper. It aims to Received in revised form 28 February 2015 achieve more accurate and reliable sensor fault diagnosis in thermal power plant. To restrain Accepted 3 March 2015 the mode mixing problem in traditional EMD, an MEEMD is proposed based on signal recon- Available online 17 March 2015 struction and pseudo component identification. The MEEMD is applied to decompose the original thermal parameter signals into a finite number of Intrinsic Mode Functions (IMFs) Keywords: and a residual to extract the sensor fault feature. After analyzing the inherent physical mean- Sensor fault diagnosis ings of each IMF and residual, the variances of them are extracted as feature eigenvectors to Empirical mode decomposition Probabilistic Neural Network