Detection of Weak Ship Radiation Characteristic Signal based on Wavelet and Chaos Array

Abstract

The ship radiated noise not only has the chaotic characteristic, but also the characteristic signal in the radiated noise is often covered by the complex marine environment noise. This makes the performance of many algorithms greatly reduced in the actual target recognition. This makes the performance of many algorithms greatly reduced in the actual target recognition. In this paper, a new detection method based on wavelet transform and chaos theory is proposed, the existence, frequency and phase of weak signal are detected by wavelet chaotic oscillator array. Firstly, a threshold selection algorithm based on wavelet packet decomposition coefficients and unbiased estimation is proposed to automatically determine the decomposition level and denoising threshold of wavelet. This method can overcome the blindness and irrationality of decomposition and threshold selection in wavelet denoising. Then, the judgment basis based on the Melnikov function and the zero-crossing detection method is proposed, which not only avoids the ambiguity between the chaotic critical state and the large-scale periodic state in the chaotic system, but also successfully detects the frequency and phase of the signal to be measured under the condition of low signal-to-noise ratio. Finally, the correlation algorithm is applied to the sea survey experiment. The experimental results show the correctness and effectiveness of the algorithm. Finally, the algorithm is used to detect the radiation noise of the actual ship, and the line spectrum component in the noise is successfully detected. This provides new theories and methods for the subsequent detection and target recognition of ships and underwater targets.