基于可观测性分析的光纤陀螺SINS/测速仪组合系统技术

FOG SINS/Speedometer integrated system technique based on observability analysis

为了简化求可观测度的过程,便于导航时利用状态变量可观测性分析结果进行姿态解算。在仔细研究了分段定常系统(PWCS)的可观测性分析方法,以及基于奇异值分解的系统状态可观测度分析方法的基础上,提出了一种改进的状态可观测性分析方法,该方法在求可观测度时,不需要量测值,利用可观测性矩阵奇异值分解后的V阵即可进行可观测性分析;利用该理论得到的结果,在SINS/测速仪组合导航解算过程中只对可观测度好的俯仰角误差和横滚角误差进行反馈校正,不反馈可观测度较差的航向角误差,仿真结果证明,这种有选择的反馈校正方法可以保证姿态角在导航开始几秒钟内就快速收敛。

In order to simplify the process of computing observable degree and use attitude algorithm to obtain the observability analysis result of state vectors,the observability analysis method of piece-wise constant system and the observable degree analysis of the system state vectors based on the singular value decomposition were studied.An improved method was proposed.The method was based on V matrix of the observability matrix after the singular value decomposition and had nothing to do with the observation vector.In the simulation of the FOG SINS/Speedometer Integrated System,the errors of pitch and roll with good observable degree are fed back based on the theory,and the error of yaw isn＇t fed back as bad observable degree.The simulation results show that this selective feedback method can ensure the errors of attitude converge quickly on the first seconds.