基于裂变自举粒子滤波的空中目标跟踪处理研究  

Air Target Tracking Processing Research Based on Fission Bootstrap Particle Filter

作　　者：李林峰[1] 张国芳[1] 王贵喜[2] 

机构地区：[1]河南工业贸易职业学院 [2]中国人民解放军92493部队

出　　处：《现代防御技术》2010年第4期142-145,共4页Modern Defence Technology

摘　　要：针对基本粒子滤波算法中的贫化现象,采用裂变自举粒子滤波算法,该算法对大权值的粒子进行裂变繁殖,覆盖粒子群中的小权值的粒子,裂变后的权值为上次粒子的均方值,保持粒子的多样性,设置粒子裂变门限防止数据量过大,最后给出空中目标模型和跟踪估计。仿真结果表明该算法的精度优于基本粒子滤波。In view of the poverty phenomenon in elementary particle filter algorithm, fission bootstrap particle filter algorithm can be adopted, which powers the large weight of the fission particles breeding, covering small weight of the particles in particle swarm. The weight value after the fission mean the square value of last participles. To maintain the diversity of particles, particle fission threshold is set to prevent data overload. Finally the target model and track the estimated air target are presented. The simulation results show that the algorithm is more accurate than the elementary particle filtering.

关 键 词：裂变自举 权值 门限 分配 

分 类 号：TN957.51[电子电信—信号与信息处理] TP391.9[电子电信—信息与通信工程]