不完全量测的变维容积卡尔曼滤波算法

Variable dimension cubature kalman filter algorithm with incomplete measurements

针对不完全量测情况下的机动目标跟踪问题,提出一种变维容积卡尔曼滤波算法。首先,根据系统状态空间模型结构采用Kalman滤波-容积Kalman滤波(KF-CKF)为基本滤波器。其次,通过计算不完全量测的一阶矩和二阶统计矩,将不完全量测滤波问题转化为确定量测滤波问题,并导出相应的状态估计方法。最后,将其与变维滤波技术相结合,提出不完全量测下的变维CKF算法。计算机仿真实验表明:新算法具有很好的估计准确度,在机动目标跟踪应用中有着良好的应用前景。

Aiming at the problem of maneuvering target tracking with incomplete measurements,a variable dimension cubature Kalman filter algorithm is proposed. Firstly, the Kalman filter-cubature Kalman filter（KF-CKF） is adopted as a basic filter according to the state space model of tracking system. Secondly,by calculating the first and second-order statistical moments of the incomplete measurements, the state filtering with incomplete measurements is converted into the state estimating with complete measurements. Then, the corresponding state estimation method is derived. Finally, combining with variable dimension filter technology, a variable dimension CKF algorithm is presented. Computer simulations show that the new algorithm has good estimation accuracy and great application prospect of maneuvering target tracking.