一种基于奇异值分解的自适应降噪方法
An adaptive noise reduction method based on singularity value decompose
-
摘要: 根据信号处理基本理论和方法,针对奇异值分解方法中有关的Hankel矩阵有效秩难以确定的难题,提出了一种奇异值分解方法,即主分量分解方法,并通过试验数据进行了验证。仿真信号和海上实录信号的降噪实验研究表明,提出的方法比基本的LMS滤波和奇异值分解降噪效果更加优越,能有效提高信噪比并去除噪声。Abstract: Based on the principal component analysis(PCA)and singular value decomposition(SVD) denoising, an improved adaptive denoising scheme is proposed. Simulation results and sea trials show that this method can eliminate the noise in the signal more effectively than the classical denoising method based on SVD. Also the waveform is recovered better and the signal-to-noise is increased markedly.