Error compensation for FOG SINS under vibration based on Elman Neural Network

To take full advantage of inertial components of fiber-optic gyroscope strapdown inertial navigation system (FOG SINS), the influence of installation error of FOG and accelerometer on the output of FOG SINS should be considered. The characteristic of vibrant signal is picked up by analyzing the signals of accelerometer with power spectral density, then an effective vibrant error compensation scheme is carried out. Method for Elman Neural Network which is fast to calculate is used to construct the model of vibrant error compensation. In the period of initial alignment, the drift error caused by vibration is compensated with the model of vibrant error compensation, and the precision of initial alignment is improved. This paper will be helpful to the future work for vibrant error compensation.