Research on the random error properties of FOG based on Total Variance

To solve the problem that the estimation error of Allan variance is great at long-term τ-values,according to the frequency properties of the FOG random errors,a method for characterizing the random errors based on Total variance is presented.The estimation performance is compared between Allan variance and Total variance with the simulated random errors.The noise types and levels of the measured FOG signal are analyzed and identified with the Total variance method.The experimental results show that,the values of Total variance are consistent with the actual power-law noise at long-term τ-values.It can efficiently improve the confidence in the case of great mean index,and its estimation performance outperforms that of Allan variance.