中文摘要：

无人机（UAV）在使用滤波器对目标跟踪时常遇到目标丢失情况,然而在目标丢失时使用一步估计代替估计值会对跟踪精度造成较大影响,鉴于此,将自适应衰减记忆滤波（AMAF）理论应用于无迹卡尔曼滤波（UKF）以提高再次捕获目标后的跟踪精度;新的滤波器命名为自适应衰减记忆卡尔曼滤波（AMAUKF）;在此基础上设计了滤波流程和计算机仿真实验,结果表明新的滤波器不但缩短了收敛步长,而且提高了跟踪精度。

英文摘要：

UAV tracking target always face the target-lost event.However,filter used to consider one-step prediction as estimation when target lost event happening,which cause bad influence to precision of tracking.For that reason,applying the Self-Adaptive Memory Attenuated Filter theory to UKF could increase the tracking precision after target re-tracked by UAV.The new filter is named as Adaptive Memory Attenuated Unscented Kalman Filter（AMAUKF）.Based on this new filter,it designs the filter running flow and computer experiment,which show the new filter shorten the step of convergence and advance the precision of tracking.