Comparative Analysis of Quaternion Based Orientation Estimators with Experimental Data from Inertial Measurement Units

In this study, different quaternion based filters for orientation estimation using inertial measurement units (IMU) were examined. Orientation determination methods using quaternion representation were detailed, and the complementary filter, Error State Kalman Filter (ESKF), Madgwick filter, Kalman based XKF3i filter, and Versatile Quaternion-based Filter (VQF) were addressed. Experimental studies were conducted using an optical motion capture system and an MTi-300 IMU, comparing the performances of different filters with the optical system taken as the reference. The results showed that all filters operated with high accuracy under a root mean square error (RMSE) below 0.2°. The VQF algorithm provided the best tracking performance of the experimental data, and its modular approach was observed to be effective in reducing error.