Study Of Scale Transformation Stochastic Resonance Of A Bistable System

Taiyong Wang,Yonggang Leng,Yonggang Xu,Wenjin Wang,Xuda Qin
2005-01-01
Abstract:Stochastic resonance (SR) has been discussed from the viewpoint of power spectrum, which presents that, Lorentz distribution of response power spectrum of a bistable system makes most power energy accumulate in lower frequency area, and that only the SR spectral spike locates at the lower frequency region can an identifiable SR phenomenon be seen. This case is in the situation of small parameters. However, with large parameters, it is impossible for the small parameter SR. For the problem, a new way of scale transformation stochastic resonance (STSR) is proposed. The proposed method can not only realize the large parameter SR-like response, but also successfully extract the weak signal from strong noise. With the large parameter STSR technique, the properties of the power spectrum of a bistable system have been thoroughly investigated, which demonstrates that, to obtain an explicit maximal SR-like spectral peak at the weak signal frequency, the ratio of sampling frequency and the signal frequency should be equal or larger than fifty. A practical engineering application, monitoring and diagnosis of mechanical faults of an electromotor, proves that the approach of large parameter STSR is of promising future in the engineering signal processing.
What problem does this paper attempt to address?