The Fast algorithm of the dynamic Allan variance for gyroscopes

The dynamic Allan variance (DAVAR) [1] is a useful tool to track non-stationaries in the behavior of gyroscopes. It allows us to diagnose whether the stability of gyroscopes is changing with time. Unfortunately, with the length of the analyzed time series increasing, the computational time of the DAVAR grows very quickly. Consequently, it costs huge time to deal with the data. In this article, a Fast algorithm of the DAVAR for gyroscopes is developed. In order to verify its validity, three sets of simulation data are characterized by the Fast algorithm and the classical one. The Fast DAVAR could reduce the computational time dramatically when the long time series is analyzed. Furthermore, a vibration experiment with fiber optic gyroscopes has been implemented to validate the good performance of the Fast algorithm. Compared with the classical one, the Fast algorithm of the DAVAR shortens the computation time 6 times. In conclusion, the Fast algorithm of the DAVAR guarantees a dramatic reduction of the computational time and outperforms the classical DAVAR.