奇异值分解故障检测法在捷联惯导中的研究

陀螺仪是惯导系统的核心元件,文章针对光纤陀螺信号的特点,将奇异值分解算法（SVD）应用于陀螺故障检测中,该方法不仅能检测出单个陀螺发生故障的情况,而且还能有效地检测出两个陀螺同时发生故障的情况,有效提高了系统的可靠性。文中以六个单自由度陀螺冗余惯性导航系统为研究对象,合理地布置测点,实现了对系统中关键的部件进行状态检测,仿真结果验证该方法是可行的。

Fault Detection of Strap-down Inertial Navigation System Based on SVD

Gyroscope is the core component of the inertial navigation system.In view of the characteristics of optical fiber gyroscope＇s signal,the Singular Value Decomposition（SVD） method was applied into the fault occurrence.The experiment proves that the SVD method can not only detect the single gyroscope fault occurrence,but also effectively detect two-gyroscope＇s simultaneous fault,which can enhance the system security.The six single-degree-of-freedom gyroscopes are studied.By reasonably arranging the detection point,the on-line detection for the state of system＇s key parts was realized.The simulation results prove that this method is feasible and can also improve the precision and reliability of the system.