A Weak Signal Detection Based On Multi-stable Stochastic Resonance

In recent years,weak signal detection based on the principle of stochastic resonance has been attracted much attention,which is widely used in many fields,such as fault diagnose,communication technology and acoustics.However,multi-stable SR method still has some problems to be solved in engineering practice.Therefore,in the paper,the multi-stable SR system is taken as system model to investigate that the multi-stable SR phenomenon excited by multiple high-frequency weak signals under complex noise.And adaptive multi-stable SR weak signal detection method based on knowledge-based particle swarm optimization(KPSO)is studied,which is used for fault diagnose of machinery.The main contents include:The dynamic equation and mathematical model of multi-stable SR is described.The switch law of that the multi-stable SR system can switch among monostable,bistable and multi-stable state is studied,and the ranges of SR system parameter when the SR system is multi-stable state are obtained.Then the mean of signal-to-noise ratio gain(MSNRI)is taken as the criterion of SR effect,the influences of system parameters(a?b?c)and noise characteristic parameters(???)on system output SR effect are studied,and the effect of different parameters on the output SR effect are revealed.The knowledge-based particle swarm optimization(KPSO)is introduced and the mean of output signal-to-noise ratio(MSNR)is taken as fitness function,which is used for optimizing multiple parameters(a?b?c)synchronously.And the parameter-induced SR is used for detecting the weak periodic signals and weak aperiodic signals under different noise underground,and the fault signals of industrial environment.The request of weak signal detection in industrial environment can be satisfied by the studied method in this paper,which lay a foundation for multi-stable SR weak signal detection application in engineering field.