GPS/INS组合导航缺星情况下的卡尔曼滤波改进算法

Improved Kalman Filtering Algorithm of GPS/INS Integrated Navigation without GPS Signal

针对GPS/INS(Global Positioning System/Inertial Navigation System)紧组合导航系统在卫星信号缺失情况下导航精度降低,容积卡尔曼滤波(Cubature Kalman Filter,CKF)应用中存在模型误差和计算误差的问题,提出了适用于该背景下的强跟踪均方根容积卡尔曼滤波(Square-Root CKF)算法。该算法通过人为地相对突出滤波过程中新数据的作用,提高了算法在模型不确定时的鲁棒性;均方根策略保证了协方差阵的正定性和对称性。仿真实验表明,改进的算法能够提高导航精度,在卫星信号缺失情况下其效能发挥地更好,提高了组合导航适应复杂环境的能力。

Aiming at the problems that the navigation accuracy of GPS/INS tightly integrated navigation was lowered under condition of GPS signal invalidation,model error and calculation error existing in the application of CKF,a strong tracking SRCKF was presented under the background.By stressing the function of new data artificially,this al-gorithm improved the robustness when the model was uncertain,and the square-root strategy ensured the positive defi-niteness and symmetry of covariance matrix.Simulation results showed that the improved algorithm could improve the navigation accuracy and had better efficiency under the condition of GPS signal invalidation.