Adaptive Invariant Extended Kalman Filter with Noise Covariance Tuning for Attitude Estimation

TL;DR

This paper introduces an adaptive quaternion-based RI-EKF for attitude estimation that uses EM algorithm for noise covariance tuning, improving stability and accuracy in sensor fusion tasks.

Contribution

It presents a novel adaptive RI-EKF method with EM-based noise covariance estimation, leveraging symmetry properties for enhanced attitude estimation.

Findings

The adaptive RI-EKF outperforms the left invariant EKF in accuracy.

The EM algorithm effectively estimates noise covariance across different window lengths.

Simulations demonstrate improved stability and convergence of the proposed method.

Abstract

Attitude estimation is crucial in aerospace engineering, robotics, and virtual reality applications, but faces difficulties due to nonlinear system dynamics and sensor limitations. This paper addresses the challenge of attitude estimation using quaternion-based adaptive right invariant extended Kalman filtering (RI-EKF) that integrates data from inertial and magnetometer sensors. Our approach applies the expectation-maximization (EM) algorithm to estimate noise covariance, exploiting RI-EKF symmetry properties. We analyze the adaptive RI-EKF's stability, convergence, and accuracy, validating its performance through simulations and comparison with the left invariant EKF. Monte Carlo simulations validate the effectiveness of our noise covariance estimation technique across various window lengths.