A Sensor Fault Diagnosis Method Research Based on Wavelet Transform and Hilbert-Huang Transform

Abstract

Focusing on the difficulty of detecting sensor fault directly in actual aerospace industrial processing we propose a kind of sensor fault diagnosis method based on wavelet transform followed with Hilbert-Huang transform. First, We detect sensor fault by continuous wavelet transform to posit fault time accurately. Then we make further analysis and processing by Hilbert-Huang transform to obtain more information about fault signals. Finally, we establish a second-order sensor model to verify control performance through Matlab/Simulink. The results of modeling and simulation show that the method is feasible and reasonable.