Robust SINS/GNSS Integration Method for High Dynamic Applications

Abstract

As high dynamic movement is always accompanied by colored noise which lacks of mathematical model, traditional Kalman filtering based on an assumption of white Gaussian noise always faces serious divergence. To enhance the performance in high dynamic environment with uncertain colored noise, a kind of robust filtering based on H-infinity technology isdeveloped. State model of the algorithm is derived from SINS error propagation. Both position and velocity errorsare used as the measurements. A simulation system which includes a tri-axial turntable and a GNSS signal simulator is used to verify the integration design under high dynamic environment. Simulation results proved that both the accuracy and robustness of the integration design have been improved significantly.

DOI: http://dx.doi.org/10.11591/telkomnika.v13i2.6876