基于无迹卡尔曼滤波的月球车惯性/天文组合导航算法研究

月球车导航传统方法主要是采用kalman滤波方法，需要对系统方程和观测方程的线性化，因此会引入的线性化误差的问题, 本文针对做出适当的改进。本文采用UKF方法作为误差状态量的最优估计方法，并以太阳敏感器观测得到的太阳高度角和太阳方位角以及测速仪的东向北向速度作为联合观测信息，将SINS和CNS所测得的月球车相关姿态和位置信息进行数据融合，估计出组合导航系统的误差状态量，从而对惯导系统的状态量进行校正，达到提高组合导航系统的导航定位精度和稳定性，减小线性化误差的目的。仿真实验证明，本文算法具有很好的位置、速度和姿态估计精度，有效的降低了线性化误差。

Research of Lunar Rover INS/CNS Navigation Method Based on Unscented Kalman Filter

he traditional lunar rover navigation estimating methods is the use of kalman filter, which is appropriate for the situation of that both of the linearized system equation and observation equation.so the linearization error will be introduced to the entire system. This paper makes appropriate improvements for this problem. The autonomous navigation method for lunar rover is based on velocity and celestial joint observation. And we use these observations to estimating the lunar rover attitude and position, which will greatly reduce the linearing error and accumulated error. Unscented kalman filter is used properly to implement optimization estimation in this method. Finally, the simulation results of SIN/CNS demonstrated the validity and feasibility of this method for lunar rover system.