自适应SICKF及在目标跟踪中的应用

针对容积卡尔曼滤波（CKF) 估计精度在系统状态或参数突变时下降的问题，结合均方根嵌入式容积卡尔曼滤波(SICKF)和强跟踪滤波(STF)思想，提出了一种自适应SICKF(ASICKF)方法。在SICKF获得高估计精度的同时引入STF条件，根据系统输出残差获得自适应渐消因子，将其引入系统输出协方差均方根阵和互协方差阵中对滤波增益进行实时修正，强迫系统输出残差序列始终正交，从而使SICKF算法具备强跟踪能力。为验证所提ASICKF算法性能，利用数值仿真将其应用于存在突变情况的目标跟踪问题中。仿真结果表明，ASICKF在系统状态突变时仍能保持较高的估计精度，算法稳定性和适应能力较好。

Adaptive SICKF and Its Application to Target Tracking

An adaptive square root imbedded cubature Kalman filter(ASICKF) method is proposed based on the square root imbedded cubature Kalman filter(SICKF) and the strong tracking filter(STF) algorithm to solve the problem that the estimation accuracy of the cubature Kalman filter decreases when the states or parameter of the system is suddenly changed.The STF condition is introduced while SICKF obtains high estimation accuracy.The adaptive fading factor is obtained according to the residual of system output which is introduced into the output covariance root mean square matrix and mutual covariance matrix so that the filter gain can be corrected in real time,and the residual sequence of system output is forced to be orthogonal,so that the ASICKF algorithm has strong tracking ability.In order to verify the performance of the proposed ASICKF algorithm,it is applied to the target tracking under the suddenly changed condition.The simulation results indicate that ASICKF can maintain high estimation accuracy when the state of the system changes suddenly,and the robustness and adaptability of the algorithm are good.