基于粒子滤波的目标跟踪抗野值算法

运用粒子滤波对目标位置进行跟踪时,测量数据的异常突变点、目标的机动转弯、粒子数量的制约和重要性密度函数的优劣都会导致估计误差较大的野值出现,这将严重影响雷达对目标的跟踪精度。现有的野值剔除方法在目标发生机动时,都存在误剔率较高的问题。针对这个问题,采用莱特准则与机动门限准则相结合的方法,提出了不确定观测点的概念,设计了一种适用于机动目标的抗野值粒子滤波算法。仿真结果表明,该方法能较好地检测和更新野值,降低跟踪误差,提高跟踪精度。

Research on Target Tracking for Outliers Rejection Based-on Particle Filter Algorithm

When Particle Filter is applied to target tracking,anomalies discontinuity in measurement data,turning maneuver of the target,the number of particles and the importance of probability density function all can lead to the occurrence of the large errors outliers,which seriously affect the target tracking. It is difficult to handle outliers,when the target maneuvers. Focusing on this problem,combining wright guidelines with maneuver threshold guidelines,this paper puts forward the concept of uncertain observation and the outliers rejection particle filter applied to maneuvering target is designed. The simulation results demonstrate that the method can effectively detect and update outliers,reduce the tracking error and improves the performance of the prediction.