A novel mechanical fault signal feature extraction method based on unsaturated piecewise tri-stable stochastic resonance. (15th January 2021)

• Record Type: Journal Article
• Authors: Zhao, Shuai; Shi, Peiming; Han, Dongying
• Is Part Of: Measurement. Volume 168(2021)
• Publication Date: 2021-01-15
• Subjects: Stochastic resonance -- Tri-stable system -- Piecewise -- Feature extraction; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2020.108374
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: A novel unsaturated piecewise tri-stable stochastic resonance (PTSR) model is proposed. A novel weak fault diagnosis method based on unsaturated piecewise tri-stable stochastic resonance. The PTSR method proposed in this article can surmount the output saturation. The effectiveness of the method is verified by simulation and example. Abstract: Stochastic resonance (SR) is widely studied in signal feature extraction. The standard tri-stable SR (STSR) method possesses superior performance than the classical bistable SR (CBSR) method in signal feature extraction, but it still has output saturation phenomenon, which will reduce the signal-to-noise ratio (SNR) of the output signal and the signal amplitude of the characteristic frequency as the CBSR method. For the purpose of avoiding the influence of output saturation on the STSR method, a piecewise tri-stable stochastic resonance (PTSR) method is proposed and applied in fault feature signal extraction. Firstly, the simulation signals are processed using the PTSR method and the STSR method separately, then the comparison shows that the output signal has larger signal amplitude and a higher SNR. Ultimately, the devised PTSR method is utilized for extracting fault characteristics of two groups of actual signals, which also has better output characteristics compared with the STSR method.