摘要
基于信号相位匹配原理,给出了利用由双平行线列阵组成的平面阵进行声源方位估计的奇异值分解算法(Singular Value Decomposition for Signal Phase Matching,SVDSPM),推导了空间谱搜索所需的时延计算公式;分析了双线阵间距对不同频率方位估计精度的影响,从而确定了双线阵的间距;分析了声源频率对方位估计精度的影响;给出了给定频率时方位估计精度与信噪比的关系并与MUSIC算法获得的结果进行了比较;给出了利用平面阵进行低空目标的声探测时存在的声探测盲区的描述方法,并分析了声源的频率、目标初始俯仰角、阵元数、阵型对声探测盲区的影响。仿真结果表明:(1)在低信噪比时SVDSPM算法的方位估计性能优于MUSIC算法;(2)声探测盲区的范围与声源的频率和目标初始俯仰角有关,而与阵元数、阵型无关。声探测盲区的分析和研究为立体阵的设计和方位估计提供了依据。
We derive the formulas for calculating time delay needed for hunting spatial spectrum and analyze the effect of the interval of parallel bilinear array on the direction of arrival （DOA） estimation accuracy, thus determining its interval. We also analyze the effect of sound source frequency on DOA estimation accuracy, especially the effect of signal to noise ratio （SNR） on DOA estimation accuracy at a given frequency. Then we compare the estimation accuracy with that obtained through using the multiple signal classification （MUSIC） algorithm. We present the method for describing the sound detection blind area which exists in detecting low altitude acoustic targets. Using the method, we analyze the effect of sound source frequency, the type of array and the number of array elements, and the initial angle of the pitch of acoustic target on the sound detection blind area. The simulation results, given in Figs. 2 through 8 in the full paper, show preliminarily that： （1） in the case of low SNR, the DOA estimation accuracy using SVDSPM algorithm is better than that using MUSIC algorithm; （2） the sound detection blind area is dependent on sound source frequency and the angle of the pitch of the acoustic target but independent of the type of array and the number of array elements.
出处
《西北工业大学学报》
EI
CAS
CSCD
北大核心
2008年第2期152-157,共6页
Journal of Northwestern Polytechnical University
基金
国家自然科学基金(60272077
60672136)资助
关键词
双平行线列阵
DOA估计
奇异值分解算法
低空目标的声探测
声探测盲区
array signal processing, signal to noise arrival （DOA） estimation, singular ratio （SNR）, parallel bilinear array, direction of value decomposition for signal phase matching （SVDSPM） algorithm, sound detection blind area
