摘要
为了从非高斯噪声环境中有效提取水下目标辐射的线谱信号,以提高水下探测系统检测水下目标的能力,提出了基于二级自适应滤波的水下目标动态谱线增强算法,该算法用基于高阶累积量能量幂函数变步长极性迭代的自适应谱线增强器作为第一级,用传统的自适应谱线增强器(ALE)作为第二级,实行串级联接。用输入信号的峭度定义了峭度信噪比,并用此分析了该算法的性能。用水下某目标辐射线谱的实测数据,对该算法的性能进行了仿真研究。结果表明:当环境噪声为瑞利噪声或混合噪声(包含均匀分布、瑞利分布和拉普拉斯分布的噪声成分)时,该算法与ALE算法相比,有良好的抑制抑高斯噪声或非高斯噪声、提高信噪比和跟踪时变信号的性能。
For greatly improving the ability of the underwater detection system to detect the underwater target, TCAFBDLE (two cascade adaptive filtering based dynamic line enhancer) algorithm is developed. In this algorithm, the higher-order cumulant-based energy power function variable step size adaptive dynamic line enhancer is used as the first cascade, whereas the traditional adaptive line enhancer(ALE) is used as the second one. The proposed algorithm is insensitive to white or colored Gaussian noise and perform well in colored non-Gaussian noise case. The kurtosis signal-to-noise ratio (KSNR) is denned by the kurtosis of the input signal. Simulation tests are dynamically conducted using the underwater moving target-radiated data. Simulation results show that the KSNR is improved about 24.4 dB in the colored Raleigh noise or 18.6 dB in the colored mixed noises composed of the uniformly distributed noise, Raleigh distributed noise and Laplacian distributed noise with respect to the original signal-to-noise ratio(SNR) and that the proposed algorithm can perform a drop of about 35.6 dB for colored Raleigh distributed noise spectrum peak or 31.7 dB for the colored mixed noise spectrum peak, whereas the ALE algorithm can only perform a drop of about 6.3dB for colored Raleigh distributed noise spectrum peak or 4.3 dB for the colored mixed noise spectrum peak.
出处
《声学学报》
EI
CSCD
北大核心
2004年第1期68-74,共7页
Acta Acustica
基金
船舶国防科技预研基金(2000J42.2.8)
西北工业大学博士创新基金(2002.04)资助项目
