A Dual Quaternion Based Fusion Framework for IMU Data with 6 DOF Pose

Abstract

Based on the highly successful application of quaternions for attitude estimation, this paper proposes an approach to position, velocity and attitude estimation for Micro Aerial Vehicles(MAVs) using dual quaternions. The states are represented in dual quaternion and time continuous states propagation model are derived via dual quaternion time update equation. At the same time, the error propagation equations based on additive error model is derived and implemented to fuse data from multiple sensors using Kalman Filter. Simulation results showed that the combination of multiple sensor data highly increase the estimate precision. In this paper, the sensor fusion algorithm is pivoted around EKF(Extended Kalman Filter) and dual quaternion.