卫星不足情况下低成本MEMS-INS/GPS伪松组合导航

为了解决GPS可观测卫星不足情况下低成本微电子机械-惯性导航系统/全球定位系统（MEMS-INS/GPS）组合导航精度维持问题，提出基于灰色模型和自适应卡尔曼滤波的MEMS-INS/GPS伪松组合导航方法。以MEMS-INS/GPS松组合导航模式为框架，建立了伪松组合导航系统的状态空间模型。基于MEMS-INS/GPS的历史观测数据，使用灰色模型对MEMSINS/GPS观测差值进行预测，称为系统伪观测量。当GPS可观测卫星充分时，使用噪声自适应估计卡尔曼滤波对MEMS-INS/GPS进行松组合导航；当GPS可观测卫星不足时，使用噪声自适应估计卡尔曼滤波依据系统伪观测量，将MEMS-INS/GPS进行伪松组合导航。以车载低成本MEMSINS/GPS组合导航系统为例进行仿真和实验验证，结果表明：当GPS可观测卫星不足时，传统的MEMS-INS/GPS松组合导航精度迅速下降并发散，而MEMS-INS/GPS伪松组合导航精度与GPS正常工作时的导航精度相差不大，维持了较高精度的导航状态。

Pseudo loosely-integrated navigation of low-cost MEMS-INS/GPS with insufficient observable satellites

To solve the problem of accuracy maintenance of low-cost MEMS-INS/GPS integrated navigation under the condition of insufficient observable GPS satellites, a MEMS-INS/GPS pseudo loosely-integrated navigation method is proposed based on grey model and adaptive Kalman filter. In the framework of the proposed navigation mode, a state space model of integrated navigation system is established. Based on the MEMS-INS/GPS historical observation data, grey model is used to predict the difference between GPS and MEMS-INS, and this prediction is named system pseudo-observation. When the observable GPS satellites are sufficient, noise adaptive estimating Kalman filter is used for MEMS-INS/GPS integrated navigation; otherwise, this filter is used for MEMS-INS/GPS pseudo loosely-integrated navigation based on the system pseudo-observation. An example of low-cost MEMS-INS/GPS integrated vehicle navigation system is used in simulation and experimental verification. The results show that when the observable GPS satellites are insufficient, the traditional MEMS-INS/GPS loosely-integrated navigation decreases in accuracy rapidly and diverges, but the MEMS-INS/GPS pseudo loosely-integrated navigation is not significantly different from the normal GPS navigation in terms of accuracy, maintaining a high precision navigation state.