论文部分内容阅读
针对非线性滤波算法在组合导航系统中的应用问题,利用泰勒级数展开对无味卡尔曼滤波(UKF)、容积卡尔曼滤波(CKF)和高斯厄米特积分滤波(GHQF)三种非线性高斯滤波算法的性能进行了比较分析;基于泰勒展开的精度分析表明,UKF和CKF从四阶项开始出现截断误差,而GHQF可以逼近任意阶精度的非线性系统的后验均值;以CNS/SAR/SINS非线性组合导航系统为应用背景,对三种滤波算法的精度进行了仿真验证。数学仿真结果表明,与UKF和CKF相比,GHQF具有更高的滤波估计精度。
Aiming at the application problem of nonlinear filtering algorithm in integrated navigation system, three kinds of nonlinear Gaussian expansion of Unscented Kalman Filter (UKF), Volumetric Kalman Filter (CKF) and Gaussian Hermitian Integral Filter (GHQF) The results of Taylor’s expansion analysis show that the UKF and CKF start truncation errors from the fourth-order term, while the GHQF can approximate the posterior mean of the non-linear system with arbitrary order precision. The CNS / SAR / SINS nonlinear integrated navigation system for the application of background, the accuracy of the three filtering algorithms were verified by simulation. Mathematical simulation results show that, compared with UKF and CKF, GHQF has higher filtering estimation accuracy.