Aerodynamic instability detection in compressor based on Hilbert-Huang transform

In this paper, Hilbert-Huang Transform (HHT), as a new time-varying signal processing method, is proposed to detect compressor aerodynamic instability. Six groups of dynamic pressure data acquired in compressor surge testes were de-noised by wavelet analysis and then processed with Hilbert-Huang transform. The result shows that it is effective to separate noise from the original signals and keep the peak signals by wavelet de-noising. By monitoring the instability frequency, early warning signals of compressor aerodynamic instability can be given through a pre-set threshold.