高斯和粒子滤波器及其在被动跟踪中的应用

为提高被动跟踪性能,提出了一种高斯和粒子滤波方法。在建立目标被动跟踪模型的基础上,使用高斯和滤波(GSF)近似目标状态的后验密度,利用粒子滤波方法处理GSF中的均值和方差计算问题,推导了高斯和粒子滤波器(GSPF)应用的具体算法步骤,使用机动目标被动跟踪仿真实例,与其它滤波器进行了仿真对比,分析了跟踪性能和RMSE误差。仿真结果表明,对于机动目标被动跟踪问题,GSPF不仅具有较高的跟踪精度,而且与一般粒子滤波器相比,GSPF具有较好的跟踪稳定性和较低的计算量。

Gaussian Sum Particle Filter for Passive Tracking

To improve the performance of passive tracking, the Gaussian sum particle filter (GSPF) was proposed. Firstly, the Gaussian sum filter (GSF) was presented to approximate the posterior density of the state based on the passive tracking model. Then, The particle filter (PF) was used to deal with the computation of means and covariances in the GSF, and the specific implementation steps of the GSPF were deduced. Finally, the tracking performance and the root-mean-square error were analyzed by maneuvering target passive tracking simulation. The results show that the GSPF not only has high tracking accuracy in passive tracking, but also has better stability and less computation amount than the PF.