A novel detection and frequency extraction method for underwater weak target based on adaptive variational mode decomposition and 3D chaotic system
摘要
Based on chaos theory, combined with adaptive variational mode decomposition (AVMD), a novel detection and frequency extraction method for underwater weak target is proposed. This study is mainly devoted to solving the problem of underwater weak target detection and frequency extraction under ultra-low signal-to-noise ratio (SNR). In this paper, a novel three-dimensional (3D) non-autonomous chaotic system with coexisting attractors and infinitely shrinking attractors is proposed, which introduces a cosine function to achieve chaos enhancement. Its dynamic behavior have been deeply analyzed through nonlinear methods, such as bifurcation diagram, Lyapunov exponent and sample entropy. Then, combined with the idea of scale transformation and geometric sequence, a detection system that can detect any frequency is designed. The system can achieve resonance with the underwater target through internal cosine term and enter large periodic or intermittent chaotic states. Considering the relationship between the generation of intermittent chaos and the frequency difference, we construct a frequency extraction method based on improved AVMD and Hilbert transform to achieve the noise reduction and envelope analysis of intermittent chaos and the frequency estimation of underwater target. The experimental results indicate that the designed detection and frequency extraction method can detect the existence of weak targets in complex marine environments and extract their true frequencies effectively. The detection SNR can reach – 49 dB, the frequency extraction error is 0.0019 Hz, and the deviation rate is less than 0.001