Ship-based Theory Of Modern Signal Radiated Noise Line Enhancement Research

Ship radiated noise is composed of continuous spectrum which has a wide range of frequency distribution and line spectrum that concentrated in the low-frequency. It contains lots of information about the ship feature, such as ship types, tonnage, speed, and the line spectrum is one of the target feature representations. Its transmission distance is long, attenuation is slow, and energy is concentrated. So using which method to enhance the line spectrum has become a research hotspot.The thesis is based on improving the signal-to-noise ratio of ship radiated noise. In order to enhance the line spectrum of ship radiated noise in the way of using modern signal theory of wavelet analysis and adaptive filtering techniques to denoicing preprocessing. To highlight the denoising effect, it proposed the LMS adaptive filtering algorithm based on wavelet threshold denoising which is combined with wavelet threshold denoising theory and LMS adaptive filtering algorithm. First, processing the ship radiated noise signal by wavelet threshold denoising to improve the signal-to-noise ratio, and then regard the processed signal as the input signal of adaptive filter to further improve the signal-to-noise ratio, to achieve the effect of ship radiated noise line spectrum enhancement.At the same time, using the wavelet multilevel transformation theory that has the characteristics of multi-resolution analysis and the Mallat fast algorithm of wavelet transform to introduce the wavelet transform into LMS adaptive filtering algorithm. It presents a new numerical analysis method of multi-scale wavelet LMS adaptive filter on the input signal of the adaptive filter for multi-scale orthogonal wavelet transform, and verifies through the test of simulation.Finally, this paper processes the different SNR of ship radiated noise simulation signal by wavelet threshold denoising method, LMS adaptive filter, wavelet threshold denoising LMS adaptive filtering algorithm, multi-scale wavelet analysis of LMS adaptive filtering algorithm separately, and contrasts the changes between the improvement of SNR and the minimum mean square error, using the power spectrum estimation Welch power signal to present the signal power spectrum chart after different algorithms processing. From the data and signal spectrum, it verifies the feasibility and efficiency of the proposed method; what’s more, ship radiated-noise signal spectrum is enhanced effectively.