Research Of Weak Signal Detection Based On Piecewise Stochastic Resonance Model

Based on the theory of stochastic resonance, this paper has designed a parameter adjustablepiecewise stochastic resonance system. Through the modeling, the calculating, the simulationexperiments and the circuit design, it is confirmed that the system can detect themedium-low-frequency weak signal in the background of strong noise, especially for themechanical fault signal. Based on the DSP technology, we have designed the hardware circuit of theself-adaptive piecewise stochastic resonance system which can automatically adjust the parameterof the system. The paper was supported by the national science fund" The research of mechanicalfault diagnosis based on the nonlinear theory in the weak signal detection and treatment technology(50875070) ".The main contents of the full paper are done as follows:Chapter one has mainly introduced the development dynamics and research status aboutstochastic resonance theory in the domestic and foreign and proposed the problem of the paper needto solve. In the chapter, we also have sure the main content of the research, the method and itsinnovation points of the paper.Chapter two generally introduces the basic theory of stochastic resonance, meanwhile itanalysis the dynamic characteristics of the typical stochastic resonance system, the reasons and theconditions of producing stochastic resonance and the measurements of using the principle ofstochastic resonance.Chapter three has proposed a piecewise stochastic resonance system based on the classicstochastic resonance model which only applicable to small parameter conditions. The detection ofthe medium-low-frequency weak signal which is common in mechanical failure prone to saturation,so it is difficult to achieve stochastic resonance in these signals detection. For the piecewise system,we have done the theory analysis, the band detection, the reasoning of SNR (Signal-to-Noise Ratio)and verification of the calculation and simulation. Through the above analysis, it’s known that thepiecewise model has better adaptability for the noise and the frequency variation.Based on the piecewise stochastic resonance, Chapter four has proposed modulation stochasticresonance aim at the detection of the medium-low-frequency weak signal which was common inmechanical failure and verified the method through calculating and simulating.Chapter five has introduced the related concepts of the adaptive stochastic resonance theory andcombining with adaptive parameter adjustment principle designed an adaptive modulate stochasticresonance system based on DSP technology. Though the analysis of the characteristics of DSP, wehave designed the hardware circuit and software system of the adaptive modulate stochastic resonance system based on TMS320F2812.Chapter six has compared the piecewise stochastic resonance which designed in this paper withthe classic weak signal detection technology, known about the superiority and insufficiency of thepiecewise stochastic resonance system and make some prospects for further research.The innovation of the paper lies in the detection of the medium-low-frequency weak signalwhich is common in mechanical failure and it is difficult to achieve stochastic resonance in thesesignals detection, putting forward a piecewise stochastic resonance model and having done therelated theory analysis and simulation test. The model has improved the detecting conditions ofonly meeting the small parameter in the adiabatic stochastic resonance theory. Combined with thepiecewise model and the sweep tracking method based on the modulation stochastic resonance, theapplication can improve the detection efficiency and extract the medium-low-frequency weak signalfrom the high noise background in the engineering environment effectively. So, it can achieve themechanical fault signal detection. Meanwhile, the paper has made the verification of calculation,programming simulation and analog circuit design. In this foundation, the paper has proposedself-adaptive detection based on DSP technology TMS320F2812 chip.