Weak Signal Detection Based On Duffing Oscillator And Its Application On DSP

Because the amplitude of weak signal is lower and it is always submerged in noise. The detection of the weak signal in the back-ground of strong noise has always been the kernel problem and full of the difficulties. Using the technology can detect the weak signal which the traditional methods can not solve (for example, weak light, small displacement,micro- vibration and so on). There are many methods to detect the weak signal at present, and the common methods include: filter method, correlation method, high step statistic method and so on. But these methods'SNR are all relative high, moreover, there still have some difficulties in processing the non-Gauss noise.A chaotic system is sensitive to regular signal but immune to noise, which makes chaos have good prospect in weak signal detection technique. The technology that combined the chaos and the weak signal detection theory can enhance the SNR, this method is effective in processing the weak signal. if we want to detect the weak signal by this method, it is necessary to make certain the bifuracation threshold value of the chaotic critical state, the common methods has low efficiency and always engender false judgement. Aiming at its deficiencies, the methods of detecting weak signal based on Duffing oscillator and a improved amplitude detecting method have been proposed in this paper.The paper has analyzed the chaos system's motion, discussed several common methods of judging the chaotic system's state. In this paper, we study on Lyapunov characteristic exponents of chaos detection system, and present a improved the method of using the exponents under the period to detect the weak signal. Using neural network, to find the relationgship between the amplitude of duffing system and the Lyapunov characteristic exponent of chaotic dynamical system, then to detect the amplitude of the weak signal. The simulation experiment result has proved that the method is more accurate, and it can lower Weak signal's SNR. In order to modify the capbility to control noise, the article combines Wavelets Transform theory and chaos detection method to detect the signal. The weak signal with strong noise processed by wavelet is taken as the input of the chaos system. At last, this paper designed a Duffing oscillator detection system based on DSP(Digital Signal Processor). The distinguish method of chaos state and signal amplitude detection method are established on this system. Laid the foundation for the practical by using the Duffing system to detect the weak signal.