Application Of Stochastic Resonance In Weak Periodical Impulse Signal Detection

In the field of mechanical fault diagnosis, weak signal detection technology is getting increasing attention. In the early stages of failure, features of the failures are usually hidden in the background noise, which are difficult to detect, thus accidents may happen with predictions remain undiscovered. Traditional detection methods tend to treat noise as a negative factor in weak signal detection. However, with further research in weak signal detection field, scientists have found that in some non-linear systems, introduction of noise would inversely increase SNR of the signal, which is called stochastic resonance. In the past two decades, stochastic resonance theory and research continues to develop, and it has an outstanding performance in the areas of weak signal detection.For the purpose of application in mechanical fault detection with stochastic resonance methods, this research analyzes the limitations of stochastic resonance in engineering applications, proposes solutions to these limitations, and uses this method in bearing faults signals analysis.This paper study includes:1. Analyzing the problems that may encounter when stochastic resonance theory is applied in engineering practice, discussing the possible influence in detection result when signal amplitude, signal frequency and noise intensity do not meet the condition of stochastic resonance respectively, and one on by one, proposing solutions to those limitations.2. Through experimental simulation method, analyzing the nonlinear response of stochastic resonance, and pointing out its impact to weak signal detection.3. Analyzing stochastic resonance phenomenon with multiple frequency signal, and pointing out its impact to weak signal detection.4. Based on the above analysis, proposing the concept of stochastic resonance preprocessing to ensure the signal could meet the limitation of stochastic resonance, so that weak signals could be detected.5. Studying the capability of stochastic resonance in detecting pulse signals, and analyzing rolling bearing fault signal characteristics. Combined with resonant demodulation method, the research successfully extracted the fault features in strong background noise, verifying the effectiveness and practicality of stochastic resonance in the field of mechanical fault diagnosis.