Detection of faults in rotating machinery using periodic time-frequency sparsity

摘要: Abstract This paper addresses the problem of extracting periodic oscillatory features in vibration signals for detecting faults rotating machinery. To extract feature, we propose an approach short-time Fourier transform (STFT) domain where feature manifests itself as a relatively sparse grid. estimate grid, formulate optimization using customized binary weights regularizer, are formulated to promote periodicity. In order solve proposed problem, develop algorithm called augmented Lagrangian majorization–minimization algorithm, which combines split shrinkage (SALSA) with (MM), and is guaranteed converge both convex non-convex formulation. As examples, applied simulated data, used tool diagnosing bearings gearboxes real compared some state-of-the-art methods. The results show that can effectively detect periodical features.