改进的辅助粒子滤波当前统计模型跟踪算法  

Current statistical model tracking algorithm based on improved auxiliary particle filter

作　　者：刘亚雷[1] 顾晓辉[1] 

机构地区：[1]南京理工大学机械工程学院,江苏南京210094

出　　处：《系统工程与电子技术》2010年第6期1206-1209,共4页Systems Engineering and Electronics

摘　　要：提出了一种改进的辅助粒子滤波算法,采用基本的常规波束形成法对声信号函数进行处理,得到目标的方位谱估计,推导了以目标方位谱函数作为重要采样密度函数的辅助粒子滤波算法。并将此改进的辅助粒子滤波算法同当前统计模型相结合,提出一种基于目标方位谱的辅助粒子滤波当前统计模型自适应跟踪滤波(current statistical model-improved auxiliary particle filter,CSM-IAPF)算法。最后通过仿真验证此方法能够减少跟踪误差,具有较好的实时性和稳定性。This paper proposes an improved auxiliary particle filter algorithm and utilizes the method of conventional beam forming（CBF） to deal with the signal function of acoustic sensors,then the spectrum bearings of the target can be estimated,which can be used as the importance sampling density function in auxiliary particle filter.So the improved current statistical model adaptive tracking algorithm is proposed based on the spectrum bearings of the target in auxiliary particle filter.Finally,it is showed through computer simulation that the algorithm has better efficiency in improving precision and real-time property.

关 键 词：纯方位目标跟踪 辅助粒子滤波 统计模型 常规波束形成法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]