一种改进自适应机动目标跟踪算法

摘要

To overcome bigger error defect of the Kalman filtering algorithm in tracking the moving tar-get at uniform velocity in“current”statistical model and unsatisfactory precision of STF in non-maneuvering segment,a new algorithm is proposed to improve the tracking precision of maneuvering segment and non-maneu-vering segment.First,the acceleration variance model based on truncation normal distribution used.Second,the prediction error covariance matrix and the fading factor are madified.Finally,the estimation error covariance matrix is calculated using the Joseph form,which is more stable and robust.The simulation results show that this algorithm has excellent performance.%针对“当前”统计模型下的卡尔曼滤波算法在跟踪匀速目标时误差较大的缺陷和强跟踪滤波器对非机动部分跟踪精度不理想的缺陷。通过改进基于截断正态分布下的加速度方差模型，提高了对非机动目标的跟踪精度；对卡尔曼滤波算法中预测误差协方差及渐消因子的计算作出修正，改进机动部分和非机动部分的精度；将目前常用的估计协方差的计算公式采用Joseph公式，增强数值的稳定性和算法的鲁棒性。仿真和实践结果表明该算法具有良好的性能。