Abstract:

Aimed at the problem of random noises in the output of data of fiber optic gyroscope (FOG), the variable step size LMS (VLMS) adaptive filtering algorithm is applied to FOG data processing. A VLMS filter is constructed on the basis of the characteristics of FOG data, which can pretreat the FOG data in real time and adjust its parameters automatically. The FOG’s performance of bias stability (BS) and angle random walk (ARW) is enhanced by using the VLMS filter. The allan variance analysis of the pre-and post-filtered FOG data shows that the VLMS algorithm is effective in disposing FOG output data, it can be used in optimizing the characteristic indexes of FOG BS by 0.019°/h and ARW by 0.001 5°/h^1/2.