基于奇异值分解自适应UKF的再入滑翔目标跟踪

Adaptive UKF based on singular value decomposition to reentry glide target tracking

针对高超声速再入滑翔飞行器(hypersonic reentry glide vehicle,HRGV)跟踪难的问题,提出了一种基于奇异值分解的自适应无迹卡尔曼滤波跟踪算法(adaptive unscented Kalman filter tracking algorithm based on singular value decomposition,SVDA-UKF)。根据此类目标的特点,首先在气动力模型基础上建立了目标状态方程,以及将目标量测量转换至东北天坐标系下建立了量测方程。其次,采用UKF算法,并在此基础上,分别通过改用间接量测更新、引入协方差矩阵的奇异值分解、设计多位自适应因子进行改进。最后,结合HRGV目标的三类滑翔轨迹进行跟踪仿真。结果表明,SVDA-UKF算法在加快计算速度的同时,还提高了滤波精度以及可靠性,实现了对HRGV目标的良好跟踪。

Aiming the problem that the hypersonic reentry glide vehicle is difficult to track,an adaptive unscented Kalman filter tracking algorithm based on singular value decomposition(SVDA-UKF)is proposed.Based on the characteristics of such goals,firstly,the target state equation is established based on the aerodynamic model.And convert target measurements to east north-up system,the measurement equation is established.Secondly,using the UKF algorithm,and on the basis,improvements are made by choose using indirect measurement update,introducing singular value decomposition of covariance matrix,and designing multiple adaptive factors.Finally,the tracking simulation is carried out combining the three types of gliding trajectories of HRGV targets.The results show that the SVDA-UKF algorithm not only accelerates the calculation speed,but also improves the filtering accuracy and reliability.The algorithm achieves good tracking of HRGV targets.