水下被动目标参数估计的群体蒙特卡罗方法  

Parameter estimation of underwater passive target using population Monte Carlo method

作　　者：陈钊[1] 王海燕[1] 刘郑国[1,2] 申晓红[1] 

机构地区：[1]西北工业大学航海学院电子与通信工程系,陕西西安710072 [2]中国船舶重工集团公司办公厅,北京100097

出　　处：《电波科学学报》2012年第6期1087-1093,1262,共8页Chinese Journal of Radio Science

基　　金：国家自然科学基金(60972153);博士点基金(20096102110038);西北工业大学基础研究基金(NPU-FFR-JC201004)

摘　　要：将一种基于时域宽带信号模型的贝叶斯高分辨方位估计方法用于水下被动目标参数估计,与传统的子空间类方法相比,该方法只需要较少的信号快拍数且仅需实数运算。同时采用群体蒙特卡罗(PMC)方法执行贝叶斯计算,能进行模型阶数和方位的联合估计,其良好的并行处理能力可显著减少程序运行时间。针对较复杂的三声源情况进行了仿真,仿真结果显示该算法给出了很好的模型阶数和方位估计性能。A Bayesian high resolution direction of arrival（DOA） estimation method based on a time-domain wideband signal model is used for underwater passive target parameter estimation. Compared to conventional subspace class wideband high resolution methods, this method needs less snapshots and only requires real arithmetic. The population Monte Carlo （PMC） method is applied for Bayesian calculation which can implement model order selection and DOA estimation simultaneously. It can reduce the time of the procedure due to its parallel processing ability. Computer simulation is implemented aimed at a slightly complex situation of three sources, simulation results show that this method gives good performance of the model order selection and DOA estimation.

关 键 词：被动目标方位估计 水下目标宽带辐射噪声 时域宽带信号模型 贝叶斯高分辨 模型阶数选择 群体蒙特卡罗 

分 类 号：TJ630.1[兵器科学与技术—武器系统与运用工程]