在线快速抑噪的自适应强跟踪滤波算法

Online fast-denoising adaptive strong tracking filter algorithm

针对强跟踪滤波算法对系统时变噪声缺乏自适应能力,导致系统状态估计精度较低的问题,提出一种可以在线估计噪声协方差阵的快速抑噪自适应强跟踪滤波算法,该算法可以抑制噪声对系统状态估计的影响,使系统状态估计迅速收敛到真实值附近.仿真实验对比了强跟踪滤波算法和快速抑噪自适应强跟踪滤波算法在噪声变化环境下的性能,结果表明:快速抑噪自适应强跟踪滤波算法具有更高的状态估计精度和自适应性.

As strong tracking filter （STF） algorithm seriously decreases the estimation accuracy of system state, owing to its little adaptability for systems with time-varying noises, a fast-denoising adaptive strong tracking filter （FASTF） algorithm with the online adaptive estimation for the noise co- variance matrixes was proposed. The effects of noises from the system state estimation were sup- pressed, and the system state estimation converged to real values quickly. Performances of STF and FASTF algorithms in environments with changing noises were compared by simulation. The experi- mental results show that FASTF algorithm has better state estimation accuracy and adaptability.