Weak Signal Detection Based On Chaos Theory

Chaos theory is a nonlinear branch of science, which is applied to different engineering field. Chaos theory is most applied in the signal detection which is currently one of the frontier subjects researched. This thesis researches chaos theory for detecting weak signal. The traditional signal detection theory processes time domain signal based on statistical theory, and the detection performance is hard to improve. The thesis analyzes the characteristics of chaos theory, and establishes two signal detection models:the chaotic detection weak periodic signal model and the transient signal detection model, which are applied to underwater target detection.Firstly, this paper studies the chaos theory and the basic characteristics of chaos, and researches the two characteristics of chaos quantity:Lyapunov exponent and fractal dimension.Secondly, the thesis establishes the chaotic detection weak periodic signal model. The model makes use of dynamics behavior characteristics of Duffing system, which is in chaos critical state. We add weak periodic signal to such a system, changing the systematic phase space state. The experiment proves that this model can detect weak periodic signals of low input SNR.Lastly, the thesis constructs the transient signal detection model based on chaos theory. Underwater signal is quite complex signal that is not totally random signal, but it has certain chaos characteristics. According to the chaotic property of underwater background noise, this model adopted single-step forecast method of radial basis function neural network, which is used to detect underwater transient signal. The experiment is proved the model for detecting air-drop model hitting water of the signals, and it verifies this model can effectively detect weak transient signals underwater noise.