Combustion Related Fualt Diagnosis of Large Diesel Engine by Analysis of Ias Based on Eemd

Large diesel engines with a high number of cylinders usually have high speed, complex crankshaft and load, which strongly complicate the diagnostics based on the instantaneous angular speed (IAS). In order to improve the discrimination of the fault features, the ensemble empirical mode decomposition (EEMD) is proposed to analyze the IAS signal. Experiments are performed on a V12 diesel engine whose injection timing is alternately 50°/70° in crank angle. First, the demodulated IAS signal is decomposed with EEMD into serial intrinsic mode functions (IMFs). Physical meaning of each IMF is explained. Then, normal and fault condition are compared by time-frequency spectrum of Hilbert-Huang transform (HHT) and FFT analysis. The results show that misfire fault can be successfully diagnosed, including identified by FFT of IMF10 and fault cylinder localization by the polar presentation method.