最小二乘滤波算法在目标跟踪中的性能比较

Application of Least Square Filtering Algorithm for Target Tracking

目标跟踪就是对目标在每个时刻的状态作实时精确的估计。根据目标在运动过程中具有轨迹连续性的特点,采用最小二乘滤波在观测数据的基础上对目标的状态进行了估计。论文对完全最小二乘滤波算法和递推增广最小二乘算法在目标跟踪的应用作了研究,通过仿真实验并对算法的跟踪性能进行了分析。结果表明:递推增广最小二乘算法的性能优于完全最小二乘滤波算法的性能。

Target tracking is a real-time and accurate estimation of the target state at every time. According to the continuity of target’s trace, the least square algorithm is used to estimate the target state based on the observational data. The paper has done a study of the total least-squares algorithm and the recursive extended least squares algorithm when applied in target tracking, and the tracking performance of these algorithms is analyzed by simulation. The simulating results prove that the performance of recursive extended least squares algorithm is superior to the total least-squares algorithm.