《Fault Diagnosis of Diesel Engine Based on EMD and AR Modes》
LU Jin-ming1,2, WANG Chun-tao2, MA Jie1
(1. School of Naval Architecture, Ocean and Civil Eng., Shanghai Jiaotong University, Shanghai 200030, China; 2. School of Mechanical and Power Eng., Jiangsu University of Science and Technology, Zhenjiang Jiangsu 212003, China)
The vibration signal of a diesel cylinder’s cover is a typical non-stationary signal. Traditional analysis of this signal in the time-frequency domain is not very effective. The interval analysis in time domain is not efficient either to realize real-time diagnosis. And it is also difficult to select the base function in wavelet transform. In this paper, the empirical mode decomposition (EMD) method is used to decompose the vibration signal into a number of intrinsic mode function (IMF) components and then the auto-regressive (AR) model of each IMF component is established. The auto-regressive parameters and the variance of remnant are regarded as the characteristic vectors and served as input parameters of SVM, and the working conditions and defaults of the diesel engine are classified. The results show that the proposed approach can classify working conditions of the diesel engine accurately, and effectively even in the case of small number of samples. And the real-time automation of the diesel engine fault diagnosis can be implemented. In order to classify the faults accurately, it is necessary to select new data as the training samples when the rotating speed of the diesel engine is changed.