“量测”恒等于0的UKF雷达系统偏差滤波算法

“Measurement”Identity 0 Based on UKF Radar System Deviation Filtering Algorithm

传统的误差配准算法在将非线性方程线性化(即在系统偏差和量测处分别进行泰勒展开,并忽略高阶微小量)时,若系统偏差较大就会引入一定的误差,使误差配准的精度随之恶化。为有效解决这一问题,提出一种“量测”恒等于0的无迹卡尔曼(Unscented Kalman Filter,UKF)雷达系统偏差滤波算法。该算法改变传统算法中伪观测量为变量的特点,将两雷达“空间几何关系”在随机量测误差处进行泰勒展开,构造出“量测”为已知值且恒等于0的非线性伪量测方程,并针对快时变系统偏差,在UKF滤波基础上融入交互式多模型(Interacting Multiple Model,IMM)的机动算法进行跟踪。仿真结果表明,相较于传统算法,该算法不仅对于恒定的系统偏差配准精度高,且能对快时变系统偏差进行实时有效估计,验证了本算法的有效性。

When the traditional error registration algorithm linearizes the nonlinear equation(that is,Taylor expansion is carried out at the system deviation and measurement points respectively,and high-order micro-quantities are ignored),if the system deviation is large,certain errors will be introduced and the accuracy of error registration will deteriorate.In order to solve this problem effectively,a method of Unscented Kalman Filter(UKF)radar system deviation filtering with“measurement”equal to 0 is proposed.This algorithm changes the characteristic of pseudo observation as variable in traditional algorithm,the two radar“spatial geometric relationships”Taylor expansion at random measurement error,constructed the“measured”as the known value and identity in the pseudo measurement equations of the 0,deviation for fast time-varying system,on the basis of the UKF filter into an interactive multiple model(IMM)tracking algorithm of mobile.Simulation results show that compared with traditional algorithms,the proposed algorithm not only has high registration accuracy,but also can effectively estimate the fast time-varying system deviation in real time,which verifies the effectiveness of the proposed algorithm.