非高斯背景下长线阵测向技术研究

发布时间：2018-06-08 03:39

本文选题：DOA估计 + 长线阵　；参考：《国防科学技术大学》2014年硕士论文

【摘要】：水下目标辐射噪声的不断降低给目标探测带来巨大挑战,迫使水下探测系统不断向大孔径长线阵的方向发展。但是在实际海洋环境中,水声信道中多号简正波的干涉效应以及非高斯环境背景噪声都会使得常规线列阵测向算法应用于长线阵时性能下降,甚至无法进行目标测向。因此开展长线阵在浅海波导和复杂环境噪声背景下的测向算法研究是其走向应用以及发展高性能水下目标探测系统的重要基础。论文基于简正波理论对浅海波导中复杂环境背景噪声下长线阵测向算法展开深入研究,主要工作和创新点如下:1、针对简正波声场下的长线阵测向模型,推导并分析了浅海波导中声场水平纵向相关性随信号频率、声源与阵列距离等参数变化的规律。在基于平面波模型的各种定向算法研究的基础上,从理论和仿真分析了浅海波导中长线阵定向出现的谱峰分离或谱峰分裂现象的原因,其研究结果表明随着声波的传播声场空间相干性会出现周期性的起伏,浅海多号简正波的叠加会造成基于平面波模型的测向算法出现谱峰偏移和分裂。2、系统研究了基于简正波模型的子空间相交算法及其改进算法,分别在不同信噪比、不同快拍数、不同信号频率等条件以及水平波矢、简正波号数估计失配、信源数目估计失配等参数失配条件下对各算法的目标定向性能进行了仿真分析,探讨了这些算法在不同浅海波导条件下长线阵测向问题的优劣和适用性。3、针对复杂海洋环境背景噪声的空间非高斯特性,利用高阶累积量可抑制高斯噪声的原理,提出了将基于简正波模型的子空间相交算法与高阶累积量相结合的算法。仿真结果结果表明:在高斯白噪声、高斯分布的空间色噪声、瑞利白噪声以及瑞利分布的空间色噪声等环境噪声情况下,该算法能够实现浅海波导多模式传播条件下长线阵的准确定向。
[Abstract]:The continuous reduction of underwater target radiated noise brings great challenge to target detection, and forces the underwater detection system to develop to the direction of large aperture linear array. However, in the actual marine environment, the interference effect of multi-signal normal wave in underwater acoustic channel and the background noise in non-Gao Si environment will make the performance of conventional linear array direction finding algorithm decline when it is applied to long linear array, and even the target direction finding can not be carried out. Therefore, the research of direction finding algorithm of long linear array in shallow water waveguide and complex ambient noise background is an important foundation for its application and the development of high performance underwater target detection system. Based on the theory of normal wave, this paper deeply studies the direction finding algorithm of long linear array in shallow water waveguide with complex background noise. The main work and innovation are as follows: 1, aiming at the direction finding model of long line array under normal wave sound field. The vertical correlation of sound field in shallow water waveguide is deduced and analyzed with the variation of signal frequency, distance between sound source and array, and so on. Based on the research of various directional algorithms based on plane wave model, the causes of spectral peak separation or peak splitting in shallow water waveguide with long linear array orientation are analyzed theoretically and numerically. The results show that the spatial coherence of sound field will fluctuate periodically with the sound wave propagation. The superposition of multiple normal waves in shallow water will result in spectral peak migration and splitting of direction-finding algorithm based on plane wave model. The subspace intersection algorithm based on normal wave model and its improved algorithm are studied systematically, respectively at different signal-to-noise ratio (SNR) and different rapid-beat number. Under the condition of different signal frequency, horizontal wave vector, normal wave number estimation mismatch and source number estimation mismatch, the target orientation performance of each algorithm is simulated and analyzed. In this paper, the advantages and disadvantages and applicability of these algorithms for long linear array direction finding under different shallow water waveguides are discussed. In view of the spatial non-Gao Si characteristics of background noise in complex marine environment, the principle of using high-order cumulants to suppress Gao Si noise is discussed. An algorithm combining subspace intersection algorithm based on normal wave model with high order cumulant is proposed. The simulation results show that under the environmental noise such as Gao Si white noise, Gao Si spatial color noise, Rayleigh white noise and Rayleigh distribution spatial color noise, etc. This algorithm can realize the accurate orientation of the long linear array under the condition of multimode propagation in shallow water waveguide.
【学位授予单位】：国防科学技术大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：P715.5

【参考文献】