论文部分内容阅读
为了在非线性、非高斯系统估计中获得更好的精度,提出一种新的unscented卡尔曼滤波(UKF).采用确定性采样方法,通过选择unscented变换中的参数α=0 85,β=2和l=0,确定出2n+1个σ点,使这些σ点完全符合非线性系统Yi=F(Xi)的高斯随机变量的均值和方差.仿真结果表明:σ点通过实际的非线性系统Yi=F(Xi)传递后,其后验均值和协方差可以达到泰勒展开式的三阶精度,广义卡尔曼滤波(EKF)只能达到一阶精度.该UKF滤波与EKF算法复杂度相近,但具有比EKF更好的估计精度.
In order to obtain better accuracy in non-linear and non-Gaussian system estimation, a new unscented Kalman filter (UKF) is proposed. By using the deterministic sampling method, by choosing unscented parameters a = 0 85 and β = 2 And l = 0, 2n + 1 sigma points are determined so that the σ points exactly match the mean and variance of the Gaussian random variables of the nonlinear system Yi = F (Xi). The simulation results show that the σ point passes the actual nonlinear system After the transfer of Yi = F (Xi), the posterior mean and covariance can reach the third order accuracy of Taylor expansion, and the generalized Kalman filter (EKF) can only achieve the first order accuracy. The UKF filter is similar in complexity to the EKF algorithm, But with better estimation accuracy than EKF.