Autonomous Regression Method for Allan Variance

An autonomous/regression-based algorithm for analyzing inertial sensor Allan variance data .


Updated 12 Mar 2018

View License

This algorithm performs the classical 5-source inertial sensor Allan variance analysis but uses non-linear regression instead of the slope method to compute the noise strength coefficients. Using non-linear regression allows the algorithm to analyze all types of sensors, even with missing sources of noise (such as quantization or rate ramp) without having to modify the code. Additionally, this method produces an actual regression model for the observed Allan variance, which can be used to draw statistical inferences on the resulting coefficients. When compared to the slope method, this algorithm produces more stable (in terms of variance) and accurate (in terms of bias) estimates of the underlying random process strengths. The provided demo script also compares results between Slope method and this algorithm (ARMAV). The provided .mat file contains sample Allan variance data from a STIM-300 IMU accelerometer and a gyroscope for demo purposes.

Cite As

Juan Jurado (2023). Autonomous Regression Method for Allan Variance (, MATLAB Central File Exchange. Retrieved .

MATLAB Release Compatibility
Created with R2017b
Compatible with any release
Platform Compatibility
Windows macOS Linux
Find more on Predictive Maintenance Toolbox in Help Center and MATLAB Answers

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!
Version Published Release Notes