Sparsity-Driven Joint Blind Deconvolution-Demodulation with Application to Motor Fault Detection

Abstract

Motor current signature analysis (MCSA) has been widely used in motor fault diagnosis by extracting characteristic frequency components in the spectrum of the stator current. However, fault signatures in the motor current are generally weak and easily influenced by noise and spectrum distortion caused by varying loads, especially in the early stage of motor faults. In this paper, we develop a sparsity-driven joint blind deconvolution-demodulation approach to extract small fault signatures of motors operating at a varying load. Results on experimental data demonstrate that our approach can effectively extract fault signatures from real noisy measurements of different load variation patterns.