陀螺仪在线性能及故障诊断AR模型和神经网络方法的研究

Research of AR Model and Neural Network on the On-line Diagnosis of Gyros

本文着眼于惯导系统中在线运行的陀螺仪性能和故障诊断的研究。针对无故障和两种故障下采集到的陀螺仪前置放大器在线输出信号 ,分别采用 AR模型法和神经网络法进行分析 ,并绘出相应的模型功率谱图。通过比较和分析之后 。

In this paper,the performance of the gyroscope's fault signal of the Inertial Navigation System(INS) has been researched.Firstly,the gyroscope's preamplifier signals are sampled under normal condition and fault condition.Secondly, the signals are analyzed by the method of AR module and Neural Network,the power spectra charts were plotted,and all the data were compared in detail.Finally we found out that the eigenvalue can be used to estimate the gyroscope's performance.