Research And Application On The Weak Signal Detection Chaos Duffing Oscillator System

The weak signal examination is always an urgent demand on the science and technology development, the goal is that the useful signal is examined from the strong noise background. For Traditional methods of detecting of weak signal, noise is the main obstacle for measuring weak signal, and the threshold input SNR is high, it is difficult to further reduce it. Therefore, although the traditional measure methods are useful, there are some difficulty in detecting weak signal,which is relative to the noise.Chaos theory will be used for signal detection, Chaos Theory is a new development in signal processing area in recent years, One of the most important is two features that the sensitivity to initial conditions for chaos and the suppression ability for noise to detecting the weak signal.Detecting weak signal by using Duffing chaotic oscillator in this paper, which mainly uses change of the phase trajectory that the system from chaotic state to periodic state of large-scale to detect weak periodic signal, this needs to determine whether the system jumps from chaotic state to a large-scale periodic state, The traditional methods are the intuitive analysis and the quantitative analysis, the intuitive analysis methods are simple but the threshold accuracy is not high, the quantitative analysis methods are high precision but are complex. This paper proposes a new method, using the nature of lebesgue measure, that it can be expressed as disjoint combination of a range of length, and is applied to the signal's time-frequency analysis based on wigner transformation. Determining threshold of the chaos system in order to distinguish the chaotic state and large-scale periodic state.For signal detection of unknown frequency and amplitude,; The paper uses Duffing chaotic system, and uses the principle that the running time of the phase trajectory and the signal under the operation cycle are the same in a large-scale periodic state, using Wigner transformation and taking maximum points corresponding to the amplitude of the frequency as the signal frequency. Requesting the amplitude of the signal in known signal frequency requirements, this paper presents a basic program and achieves the more effective results by the experiment simulation, and further improves the detection accuracy and SNR.