Research Of Weak Signal Detection System Based On Adaptive Stochastic Resonance

In the field of precision measurement, signals tend to be drowned orcoveraged by noise, how to pick up the weak signal information under the noisebackground exactly has been a hotpot in recent years. Stochastic Resonance is a newtype of weak signal detection method, it is different from the methods in weak signaldetection for the purpose of suppressing noise, stochastic resonance is using noise tostrengthen the weak signal and there exists a mechanism of transferring noise energyto signal energy.In the practical application of stochastic resonance, there exists a problem that itis hard to control the appearance of stochastic resonance. Producing stochasticresonance needs strict conditions, it can be produced only when three ingredientswhich are weak periodic signal, noise and nonlinear system achieve some strictmatching relations. In the actual signal detection process, weak periodic signal, noiseordinarily can`t satisfy the condition of producing stochastic resonance. Under thepremise of without changing the weak periodic signal and noise, to purposelyproduce or make stochastic resonance more intensity, we can adjust the nolinearsystem parmeters adaptively. To achieve the goal, the paper chosed the nonlinearbistable system stochastic resonance as the research object, this paper developed aweak signal detection system, which used AVR monolithic microprocessor to achieveadaptive stochastic resonance through linear random search algorithm to drive andcontrol the digital potentiometer AD7376to change the structure parameters of thebistable system, so as to realize the weak signal detection system based on adaptivestochastic resonance. The paper did the following researches:Firstly, the paper studied the influence of the bistable system’s parmeters tostochastic resonance, studied the influence of the bistable system’s parameters tosystem barrier、system transition、system response speed and the influence of thebistable system’s optimum parameters to stochastic resonance. Through the study ofthe influence of bistable system to stochastic resonance, providing a theoreticalguidance for us to realize stochastic resonance by controlling the parmeters of thenolinear system. Secondly, the paper studied the stochastic resonance characteristic of thebistable circuit, a simulation experiment was done in the Multisim, the experimentverified the bistable circuit can well realize the function of stochastic resonance,providing the strong support for the actual bistable circuit production.Thirdly, the Gaussian white noise source was designed, the hardware circuit wasdesigned for the system, including AD7376bistable stochastic resonance circuit, A/Dswitching circuit, voltage polarity switching circuit, the power supply circuit. Thecharacteristics of bistable stochastic resonance circuit and the realization of multisimsimulation were analyzed. And the experimental device based on orifice plate wasdesigned for the system testing.Then the core control part was completed in this paper using C language andmodular design method based on the AVR microcomputer, which included the A/Dacquisition module, the driver test design of AD7376, the algorithm was designedbased on the SNR adaptive stochastic resonance, the achievement of decisionalgorithm was finished based on the statistical mean value and the mean variationadaptive stochastic resonance system.Finally, the adaptive weak signal detection system was tested for some relatedperformance in the environment of the laboratory. Experimental test results show thatfor a single low frequency (within100Hz) weak periodic signal under a certainvibration noise, the system can achieve a better adaptive weak signals’detection.