基于AR模型的非线性目标跟踪自适应算法

针对Jerk模型中各参数设置不合理对跟踪系统所造成的影响,提出一种基于自回归（AR）模型的Jerk参数自适应改进算法,实时估计并调整系统的参数,提高系统的跟踪精度及稳定性;同时,针对非线性目标跟踪系统扩展卡尔曼滤波算法（EKF）计算复杂跟踪精度低,提出采用平方根容积卡尔曼滤波器（SRCKF）进行状态估计,保证跟踪系统的精度和鲁棒性,为Jerk模型参数自适应提供良好条件.仿真结果验证了算法的有效性.

Adaptive algorithm of nonlinear target tracking based on AR model

Due to the influence of setting up the unreasonable parameters about Jerk model,an improved adaptive algorithm of parameters based on auto regressive(AR) model was proposed in this paper.The model parameters were estimated online and then the target tracking system was amended.The accuracy and stability could be enhanced effectively.Meanwhile,in order to deal with the problems of extended Kalman filter(EKF) that it was complexly computed with low accuracy in state estimation,an improved filter square-root cubature Kalman filter(SRCKF) was present,which enhanced the algorithm numerical stability,guaranteed positive semi-definiteness of the state covariance,and also increased the filtering accuracy,providing a fitness advantage for Jerk parameters adaptation.At last,simulation results verified the effective of this algorithm.