基于模态分解的最小方差模态滤波器  

A New and Better Method of MVDR(Minimum Variance Distortionless Response) Mode Filtering Based on Mode Decomposition

作　　者：易锋[1] 孙超[1] 向龙凤[1] 刘宗伟[1] 

机构地区：[1]西北工业大学航海学院,陕西西安710072

出　　处：《西北工业大学学报》2012年第1期112-116,共5页Journal of Northwestern Polytechnical University

基　　金：国家安全重大基础研究(613110020101);国家自然科学基金(60872146)资助

摘　　要：针对浅海的复杂水声环境,提出了一种基于模态分解的最小方差无失真响应(MVDR)的自适应模态滤波器,利用接收阵列的声场数据和声速梯度数据,估计了各阶简正波复幅度系数。最小方差模态滤波器提取声场某阶模态时,可以抑制其它模态和噪声的影响,减弱模态滤波器对阵列长度的依赖性及对噪声的灵敏度。通过计算机仿真研究了最小方差模态滤波器的性能,并与其它线性模态滤波器进行了对比,结果表明:该自适应模态滤波器比其它线性模态滤波器能更精确地估计各阶简正波复幅度系数,并且具有良好的环境适用性特点。depend on and better Mode filters are commonly implemented using the sampled mode shape or pseudoinverse filters, but they the received array length and have sensitivity to white noise. We propose what we believe to be a new adaptive mode filter based on the MVDR filtering. We explain it in sections 1,2 and 3 of the full paper. Their core consists of： （ 1 ） the MVDR mode filter can estimate the complex mode coefficients by suppressing other modes and noise; （2） the performance of the MVDR mode filter is analyzed by computer simulation based on the received data and the sound velocity profile in shallow water. The simulation results, presented in Figs. 2, 3 and 4, and their analysis show preliminarily that ： （ 1 ） the MVDR mode filter achieves the same performance as the reduced rank pseudo-inverse mode filter and offers greater performance than other linear mode filters; （2） the MVDR mode filter can estimate complex mode coefficients more accurately than other linear filters and adapt the filter coefficients to the changing environments using the received data.

关 键 词：自适应滤波器 模态滤波器 模态分离 浅海 

分 类 号：O042.71[理学]