针对系统模型的特点,对平方根UKF算法进行了改进.改进后的平方根UKF与标准的平方根UKF相比,在保证状态方差非负定特性的同时,减少了滤波的计算复杂度.为了提高导航系统对姿态测量性能,针对姿态误差较大时线性导航误差方程无法使用的问题,推导出适用于大姿态误差角时的地磁姿态量测方程,并结合改进的平方根UKF算法设计了非线性地磁/GPS/SINS组合导航算法.利用Matlab对大姿态误差角时的非线性地磁/GPS/SINS组合导航系统进行了仿真.仿真结果表明,通过引入地磁姿态测量信息,非线性地磁/GPS/SINS组合导航系统可以直接观测载体的平台姿态误差,进而增强了组合导航系统对姿态误差的估计效果,且可以完成大姿态误差下的导航任务.
The square-root UKF algorithm was improved according to the peculiarity of the system model.Compared with the old square-root UKF,the improved square-root UKF could not only guarantee the positive semi-definiteness of the state covariance,but could also reduce the computational complexity.To overcome the issue of linear navigation errors,equations could not be applied in the case of a system with large attitude errors.To improve the attitude determination performance of the navigation system,a nonlinear geomagnetic attitude measurement equation was derived which could be applied when the attitude errors were large.A nonlinear geomagnetic/GPS/SINS integrated navigation algorithm was proposed in combination with the improved square-root UKF.The performance of a nonlinear geomagnetic/GPS/SINS navigation system was studied by Matlab simulating when the attitude errors were large.The simulation results show that the nonlinear geomagnetic/GPS/SINS navigation system can observe the carrier platform attitude error directly through introducing the information of geomagnetic attitude measurement.As a result,the estimated effect of attitude errors is enhanced,and the navigation task with large attitude errors can be completed.