Computation of Extended Robust Kalman Filter for Real-Time Attitude and Position Estimation
This work addresses the need for robust real-time state estimation in autonomous vehicles, but the contribution is incremental as it applies existing methods to a specific domain.
The paper implements the extended robust Kalman filter (ERKF) for real-time attitude and position estimation, using QR decomposition with Givens rotation for parallel computation. The filter is applied to a cargo transport vehicle, demonstrating its real-time feasibility.
This paper deals with the implementation of the extended robust Kalman filter (ERKF) which was developed considering uncertainties in the parameter matrices of the underlying state-space model. A key contribution of this work is the demonstration of a method for real-time computation of the filter on parallel computing devices. The solution of the filter is expressed as a set of simultaneous linear equations, which can then be evaluated based on QR decomposition using Givens rotation. This paper also presents the application of the ERKF in the development of an attitude and position reference system for a cargo transport vehicle. This work concludes by analyzing the performance of the ERKF and verifying the validity of the Givens rotation method.