在GPS/MEMS-INS( Micro-electromechanical System-Inertial Navigation System)组合导航中，当载体处于恶劣环境下或者载体处于大机动运动情况下，会导致GPS失锁。此时单独工作的低精度MEMS-INS会由于其位置和速度误差随着时间的变化而迅速积累最终无法导航。针对此问题，设计了一种结构简单易于实现的神经网络辅助的GPS/MEMS-INS组合导航系统。通过模拟实验，与标准卡尔曼滤波( KF)和无迹卡尔曼滤波( UKF)框架下组合导航相比，所提出的结构简单容易实现的神经网络辅助的组合导航系统具有较高的稳定性，并且组合导航整体过程中经纬度与速度精度均提高了65％左右。%In the GPS/MEMS-INS integrated navigation GPS signals were prone to lose when the carrier was in harsh environment or in high-speed movement. In this situation the position and velocity errors of the MEMS-INS of low precision were accumulated quickly and finally the MEMS-INS completely lost its navigation ability. For the demerit a simple and easy to implement GPS/MEMS-INS integrated navigation system was developed assisting by neural network. Compared with the traditional integrated navigation algorithm based on the standard Kalman filter and unscented Kalman filter the proposed method received higher stability through the simulation experiment. Mo-reover the accuracy of location and velocity is increased approximately by 65% in the whole process of integrated navigation.
展开▼
