摘要
该文研究了杂波干扰下适用于平面机动目标实时定位的粒子滤波器的设计和实现。运用贝叶斯递推方法,分析了通用粒子滤波算法。应用马尔可夫链理论,建立了目标机动的多模式模型。针对约束仅方位跟踪(CBOT)的实际条件,推导了粒子滤波器中的最优重要密度公式。综合粒子滤波算法和约束仅方位跟踪理论,得出了约束仅方位跟踪粒子滤波器的伪码。对约束仅方位跟踪粒子滤波器的研究表明:在目标受到道路图之类的约束假设下,可以去除对本站作特定机动的要求;可在杂波环境中、目标作任意非线性机动的情况下,实现仅方位跟踪;跟踪定位距离大、角度宽;采用最优重要密度,可大大减少粒子数,降低计算量。
This paper studies the design and implementation method of particle filters suitable for the plane maneuvering target＇ s real time tracking. A general particle algorithm is analyzed by using Bayesian recursive method and a multi-mode model of target tracking is established by using Markov chain method. Aiming at the constrained bearing only tracking （CBOT） ＇ s actual condition, the formula of optimal importance density （OlD） used in particle filters is derived. The pseudo code integrating the particle filter algorithm and CBOT theory is obtained. The research and simulation results show the following conclusion： this algorithm avoids the own station＇ s maneuvering condition when the target motion subjects to some kinds of constraints; it can track arbitrary maneuvering of the target in clutter environment; its location scale and angle are more than the traditional tracking＇ s; it reduces the particle number and calculation complexity by OID.
出处
《南京理工大学学报》
EI
CAS
CSCD
北大核心
2008年第6期767-771,共5页
Journal of Nanjing University of Science and Technology
关键词
目标跟踪
贝叶斯方法
约束仅方位跟踪
粒子滤波器
最优重要密度
target tracking
Bayesian method
constrained bearing only tracking
particle filters
optimal importance density
