一种新型变增益UKF

A New Type of Variable Gain UKF

UKF是一种适应用于非线性系统的非线性滤波器,其没有进行Jaclbian矩阵计算,有着EKF无法比较的优点。由于系统建模不准确等原因,UKF滤波算法不具备对系统突变状态的跟踪能力,使得在系统出现扰动或观测异常时会出现不稳定甚至发散等现象。提出了一种新型变增益无迹kalman滤波,并与标准无迹kalman滤波器进行了仿真实验对比。实验结果显示,该滤波器具有对系统突一次的强跟踪能力。

The unscented kalman filter is a nonlinear filter for nonlinear system.There is no Jaclbian matrix calculation,it has the advantages that EKF cannot compare.Because of inaccuracies of system modeling and others,UKF filtering algorithm does not have the ability to track the mutation state of system and the system is likely to become unstable even divergent when the disturbance or observation anomaly appear.This paper puts forward a new kind of variable gain unscented kalman filter and compares it with standard UKF by simulation experiment.