学术论文

      矢量声纳高速运动目标稳健高分辨方位估计*

      A high resolution robust localization approach of high speed target based on vector sonar?

      摘要:
        针对水声矢量信号处理框架中的高速运动目标低信噪比小快拍条件下的稳健高分辨方位估计问题,将压缩感知技术应用于水声矢量信号空间谱估计模型中。结合声矢量传感器结构特性,探讨了基于声压振速联合处理的广义时域滤波方法;结合矩阵空域预滤波理论,设计了基于阻带约束通带均方误差最大值最小的空域滤波器,研究了矢量声纳空域预滤波方法;结合以上分析,提出了基于压缩感知技术的时空联合滤波高分辨方位估计方法,给出了方法的数学模型、物理解释及具体实施步骤。理论分析和计算机仿真试验表明,新方法对于小快拍数条件下的矢量声纳高速运动目标高分辨方位估计问题,具有较低的双目标分辨门限和较高的估计精度,有着良好的应用前景。湖上试验验证了方法的有效性。
      Abstract:
      Based on high speed moving target robust high-resolution direction of arrival (DOA) estimation problem under low signal-to-noise ratio and a small number of snapshots in the underwater acoustic vector signal processing framework, a novel spatial spectrum model combined with compressive sensing method is proposed. By studying the acoustic vector sensor structure, a generalized temporal filtering method based on sound pressure and particle velocity combined treatment is presented. According to the matrix spatial prefiltering theory, a new spatial filter with stopband constraint and mean square error min-max principle in passband is proposed which is used as vector sonar spatial prefiltering algorithm. Based on the methods above, a novel time-space domain jointly filtering high-resolution DOA estimation algorithm based on compressive sensing is proposed. The mathematical model, physical interpretation, and specific implement are explained in detail. Theoretical analysis and computer simulation results show that the new method has a lower dual-target distinguishing threshold and a higher estimation accuracy in solving the vector sonar high speed moving target robust DOA estimation problem under a small number of snapshots (single snapshot) condition. The higher robustness and better results of the proposed method are verified in the lake experiment.
      Author: Liang Guo-Long Ma Wei Fan Zhan Wang Yi-Lin
      作者单位: 哈尔滨工程大学,水声技术重点实验室,哈尔滨 150001
      刊 名: 物理学报 ISTICSCIPKU
      Journal: Acta Physica Sinica
      年,卷(期): 2013, (14)
      机标分类号: TB5 TN9
      在线出版日期: 2013年8月3日
      基金项目: 国家自然科学基金(批准号:51279043)资助的课题.*Project supported by the National Natural Science Foundation of China