一种Sage-Husa和可观测度的滤波算法研究

Research on a Filtering Algorithm Based on Sage-Husa and Observable Degree

针对目标跟踪中过程噪声统计特性未知和状态分量可观测度差而导致滤波精度不高甚至滤波发散的问题,提出了一种复合自适应滤波算法。我该算法在滤波过程中,利用Sage-Husa噪声估计器在线估计过程噪声,用可观测度分析方法抑制状态分量可观测度差对滤波器的不良影响。在滤波过程中实时估计和修正过程噪声的统计特性,同时对观测度差的分量通道进行滤波增益衰减,以减小状态估计误差,提高滤波算法的估计精度。解决了一类过程噪声统计特性未知且系统状态分量可观测度差的状态估计问题。仿真结果显示,提出的复合自适应滤波算法对比传统Sage-Husa滤波和可观测度分析方法能够抑制过程噪声时变和状态分量可观测度差对滤波器的不良影响,具有更高的估计精度。

Aiming at the problem that the statistical characteristics of the process noise in the target tracking are unknown and the observable degree is poor,the filtering accuracy is not high and even the filtering is divergent.A hybrid adaptive filtering algorithm is proposed.In the filtering process,the Sage-Husa noise estimator is used to estimate the process noise online,and the observable degree analysis method is used to suppress the adverse effect of the observable degree poor on the filter.In the filtering process,the statistical characteristics of the process noise are estimated and corrected in real time,and the filter gain attenuation is performed on the component channel with poor observation degree to reduce the state estimation error and improve the estimation accuracy of the filtering algorithm.A class of state estimation problems with unknown statistical characteristics of process noise and poor observable degree of system state components.The simulation results show that the proposed composite adaptive filtering algorithm can suppress the adverse effects of process noise time-varying and observable degree poor on the filter compared with the traditional Sage-Husa filtering and observable degree analysis method,and has higher estimation accuracy.