A Comparative Analysis of Sensor Fusion Algorithms for Miniature IMU Measurements

Abstract

Inertial measurement units, typically consisting of tri-axis gyroscopes and accelerometers, are very important for a plethora of applications in the upcoming Tactile Internet. Yet, especially for miniature devices relying on cheap electronics, their measurements are often inaccurate and subject to gyroscope drift, which implies the necessity for sensor fusion algorithms. These are critical for estimating the orientation of an object by combining multiple measurements, and also require fast computation to be useful in practice. This paper presents a comparative analysis of a standard trigonometry computation, shown to be ineffective, with popular candidate algorithms, namely, Kalman, Mahony, and Madgwick, with a specific focus on their suitability for small embedded systems. The algorithms were evaluated on experimental data based on their accuracy and computational efficiency.