针对高速自治水下航行器的UKF主动目标跟踪算法

A UKF-based Active Target Tracking Algorithm for High-speed Autonomous Underwater Vehicles

研究了基于高速自治水下航行器平台下的主动单目标跟踪,基于Unscented Kalman Filter(UKF)建立跟踪滤波器,在强观测噪声、大采样时间间隔情况下完成对目标各运动状态参量的准确估计。将此跟踪滤波器与基于Extended Kalman Filter(EKF)的跟踪滤波器进行了对比。计算机仿真结果表明采用EKF滤波器,目标的速度估计值可以收敛向真值,而距离估计值无法获得收敛;采用UKF滤波器,目标的速度和距离估计值都能获得收敛,且其对目标的速度估计较EKF准确。

The problem of active tracking for a unitary target based on a platform of high-speed autonomous underwater vehicles （AUV）, was researched. A robust Unscented Kalman Filter （UKF） based tracking algorithm was founded. In case of strong observation noises and long sampling intervals, it led to estimate for the state parameters, which were used to describe the target＇s movement real time. This filter was also compared with Extended Kalman Filter （EKF） for this application. Simulation results show that the EKF can lead to some accurate estimate for the target＇s speed values, meanwhile, the estimates for the target＇s distance values become divergent. Using the UKF, both of the speed and distance values can be estimated accurately. Moreover, the speed values can be estimated more accurately than using EKF.