Research on Multi-Frequency Weak Signal Detection Based on Adaptive Flexible Stochastic Resonance

Abstract

According to the problem that multi-frequency weak signal under the full frequency coverage constraint is hard to detect, this paper puts forward a detection method by adaptive flexible stochastic resonance. Based on frequency band decomposition of input signal, each function modules of flexible stochastic resonance system are configured for different frequency bands. Parameters including system structural parameters, frequency compression factor and frequency of modulation signal are optimized adaptively by genetic algorithm. Then each signal component distributed on each frequency band is processed by stochastic resonance effect to realize the enhancement of multiple weak features under the full frequency coverage constraint. Finally, through simulation analysis and an example of rolling bearing fault diagnosis, the effectiveness and comprehensiveness of this presented method in the detection of multi-frequency weak signal is verified.