一种基于粒子滤波的概率数据关联方法

A Possibility Data Association Algorithm based on Particle Filtering

针对复杂环境下的机动目标跟踪问题,提出了一种基于粒子滤波的概率数据关联方法(PF-PDA),首先利用跟踪门对回波进行预处理,筛选出有效回波;随后利用粒子滤波,对关联概率中的残差协方差阵进行修正,求得关联概率,进而得到融合测量;最后对目标状态进行更新。与PDA的仿真比较表明,在满足系统实时性的前提下,本方法在跟踪精度上有很大提高。

For the problem of maneuvering target tracking in complex environment, a probability data association method based on particle filtering is proposed. The raw measurements are pre-processed firstly with tracking gate and valid measurements are selected.The particle filtering is invoked to modify the residual covariance matrix and to get the fusion value of valid measurements,which is used to update the target state.Compared with the PDA algorithm in simulation,the presented technique improves the tracking precision with acceptable system real time requirement.