Investigation on Reciprocating Engine Pattern Recognition via Improved Hilbert Spectrum

摘要

Hilbert spectrum (HS) is developed according to instantaneous frequency analysis by using a new kind of signal analysis method called empirical mode decomposition (EMD). The original signal can be decomposed into several IMFs and a residual component via EMD process. But some intrinsic mode functions are less meaningful and illusive component. To improve the performance of EMD process and the corresponding HS, a improved method have been investigated in this research. Noise and illusive components are removed and can improve the accuracy of Hilbert spectrum. An experiment of marine diesel engine pattern recognition is put forward to testify the effectiveness of this method.