一种基于多重次自适应因子的强跟踪UKF算法

An algorithm based on multiple adaptive factors strong tracking UKF

针对强跟踪UKF算法存在自适应因子调节单一、难以准确调节多维的系统变量对不准确系统模型和噪声自适应调节能力不够的问题,提出了一种基于多重次自适应因子的强跟踪UKF(MST-UKF)算法.首先解析了强跟踪UKF的本质原理;其次提出在状态噪声协方差和观测噪声协方差前各引入多重次自适应因子对角矩阵对其进行自适应调节,并对其计算方法进行了设计;最后分别针对系统模型、噪声不准确情况下的目标进行跟踪仿真.仿真结果表明,MST-UKF算法能自适应直接调节噪声协方差,以此应对系统模型的不匹配、状态噪声的不准确性以及观测噪声的不准确性,实现了对复杂条件下目标的良好跟踪.

In the strong tracking unscented Kalman filter(UKF)algorithm,the simplex regulation of adaptive factors has difficulty in accurately adjusting multi-dimensional system variables,and also has insufficient adaptive adjustment ability for inaccurate system models and noise.In order to solve the above problems,this paper proposes an algorithm based on multiple adaptive factors strong tracking unscented Kalman filter(MST-UKF).Firstly,the essential principle of strong tracking UKF is analyzed.Secondly,multiple adaptive factor diagonal matrix is introduced before the state noise covariance and observation noise covariance respectively to adjust them adaptively,and the calculation method is also designed.Finally,tracking simulation is carried out for the targets under the condition of inaccurate system model and inaccurate noise,respectively.The simulation results show that the MST-UKF algorithm can directly adjust the noise covariance adaptively to deal with the mismatch of the system model,the inaccuracy of the state noise and the inaccuracy of the observation noise,thus achieving a good tracking of the targets under complex conditions.