On-line modeling and real-time filtering of the FOG's random drift
Yang Mengxing · Optical Technique · 2011
For the random error of the high-precise Fiber Optic Gyro(FOG),it can be found that its time-relativity is weak based on the analysis of the interrelated-function of time series.Then the wrong model is identified easily by adopted the traditional way.In this situation,norm of AIC is used to determine the best step of AR-model and ARMA-model,and then Kalman filter is used to filter the random drift.At last,the mathematic model is determined by the bias's stability improvement of the filtered outing.Verification result shows that both AR-model and ARMA-model are suitably used for the mathematic model of FOG's random drift and can pass model applicability test,and ARMA-model is more suitable.