Study On Parameters Optimization Of Stochastic Resonance And Its Application

Stochastic resonance is the nonlinear phenomenon, which is caused by thesynergistic effect from weak periodic signal and nonlinear system with the mediumof noise, involved in the parameters of periodic signal amplitude, frequency, noiseintensitiy and nonlinear system parameters. In practical applications, the input signaland noise is given and their parameters can not be changed. Only by adjusting theparameters of the nonlinear bistable system to make the bistable system match withthe input signal, the system can generate stochastic resonance. After the influence ofthe bistable system parameters on stochastic resonance analyzed and thecharacteristics of the adaptive algorithm studied, the adaptive stochastic resonancebased on artificial fish algorithm is put forward.The influence of the bistable system parameters on the barrier height and thechange of the system output SNR along with bistable system parameters are analyzed.By adjusting the bistable system parameters, it can realize the generating andenhancing of stochasic resonance.The characteristic of the commonly used adaptive algorithm in the practicalapplication are analyzed. Linear stochastic search algorithm adopts the method ofweight superposition, resulting in that it is not sure to get the global optimal solution.Because of the introduction of random mutation, genetic algorithm is used tosearching to the wrong space. Therefore, the adaptive stochastic resonance based onartificial fish algorithm is put forward. Artificial fish swarm algorithm is uesd toadaptively adjust the parameters of the bistable system, in order to realize thegenerating stochastic resonance. Through nonlinear coupling mode, the couplingsystem is formed from two bistable system. A feedback method is proposed tocontrol the generation of stochastic resonance with coupling, and then enhanceresonance effect with the optimization of control parameters.The adaptive stochastic resonance based on the artificial fish swarm algorithm is applied to bearing rolling element fault detection, bearing inner ring fault detectionand vortex signal detection in different flow. The fault characteristic frequency andthe frequency of vortex street are obtained successfully. The experimental resultsshow that by using artificial fish algorithm to optimize bistable system parametersparallelly, it can enhance the weak signal, and then improve signal-to-noise ratio,therefore, weak signal detection is realized effectively.Finally, intelligent system uesd to detect weak signal has been developed withthe seamless integration of LabVIEW and MATLAB based on COM technology.According to the characteristics of different measured signals, the intelligentdetection systerm can adaptively adjust the bistable system parameters to realizestochastic resonance. The detection of vortex signal shows that the system can realizeweak characteristic signal detection effectively and has broad application prospect.