GFSINS姿态角速度双路组合方案设计

针对无陀螺捷联惯导系统解算载体姿态角速度精度不高的系统瓶颈,分析现有的一种九加速度计配置方案角速度信息输出特征,提出通过改进解算方法,在载体上同时获取两套相对完整而独立的姿态角速度信息从而构成姿态角速度双路组合方案的设想,并且在该组合方案中引入反馈型自适应神经网络,通过设置合理的神经网络学习周期,提供更逼近真实值的加速度计输出,结合包含加速度计误差少、没有加速度计误差积累两种优点的姿态角速度辅助算法,以获得更高精度的载体姿态角速度输出,进一步提高无陀螺捷联惯导系统导航精度.仿真结果验证了该设计思想可行性,并且证明了角速度双路组合方案相对普通角速度解算方案在精度上的优势.

Design of dual attitude angular-rates combined scheme in GFSINS

In view of the precision limitation in processing attitude angular-rate in GFSINS,the characteristics of attitude angular-rate output from a nine-accelerometer disposition scheme were analyzed,and a combined scheme of dual attitude angular-rates based on this accelerometer disposition scheme was put forward.A feedback adaptive neural network was given in the combined scheme to approach the true acceleration outputs in a proper learning period.With these compensated accelerometer outputs,the attitude angular-rate aided algorithm was designed based on the combined scheme,which not only introduce less accelerometer error but also avoid accelerometer error accumulation.The feasibility and precision superiority of the combined scheme were proved by the Matlab simulation results.