A time‐controllable Allan variance method for MEMS IMU

Purpose The purpose of this paper is to reduce the calculation burden and speed up the estimation process of Allan variance method while ensuring the exactness of the analysis results. Design/methodology/approach A series of six‐hour static tests have been implemented at room temperature, and the static measurements have been collected from MEMS IMU. In order to characterize the various types of random noise terms for the IMU, the basic definition and main procedure of the Allan variance method are investigated. Unlike the normal Allan variance method, which has the shortcomings of processing large data sets and requiring long computation time, a modified Allan variance method is proposed based on the features of data distribution in the log‐log plot of the Allan standard deviation versus the averaging time. Findings Experiment results demonstrate that the modified Allan variance method can effectively estimate the noise coefficients for MEMS IMU, with controllable computation time and acceptable estimation accuracy. Originality/value This paper proposes a time‐controllable Allan variance method which can quickly and accurately identify different noise terms imposed by the stochastic fluctuations.