Allan variance the stability analysis algorithm for MEMS based inertial sensors stochastic error

MEMS based inertial sensors are widely used due to their small size, low cost, and low power requirements. Inertial sensors are graded as per the error exhibited by them, therefore for any application at hand the error model of these sensors is explicitly considered in the unit model. In this paper, post calibration residual and stochastic errors are modelled by time domain stability analysis standard, the Allan variance (AV). Cluster sampling based various variance techniques with improvement in estimation accuracy and confidence of interval are considered. The effective degree of freedom for overlapping AV, modified AV and total variance techniques are calculated with chi-square statistic. Temperature effect on AV is observed and stochastic error coefficients are extracted from experimental data for error model of inertial sensors. The reported results are within 1s confidence of interval of inertial sensors specification´s datasheet provided by the manufacturer.