A Nonlinear Observer for Integration of GPS and INS Attitude

Bjørnar Vikand Thor I.Fossen

Abstract:	The dynamics that describe rotation between two refer-ence frames are in general nonlinear. Consequently, most, if not all, approaches to integration of INS and GPS at-titude use an Extended Kalman ¿lter (EKF) to perform the integration. The EKF performs well in most practi-cal applications, but it is an approximate ¿lter and diver-gence may occur from that reason. This paper presents an observer which has been designed using a nonlinear control theoretic approach. GPS measurements are used to estimate gyro biases, scale factor errors and misalign-ments. When the estimation errors are small, the observer becomes in effect a Luenberger observer. In this case, if Kalman ¿lter gains are used, minimum variance esti-mates will be obtained. The observer has been proven to be exponentially convergent, that is, the convergence properties are essentially the same as for a linear system. Robust stability in the presence of noise and any other bounded disturbances has also been shown. The observer has been compared to the EKF using actual GPS and IMU attitude data from a marine application, and the two show similar performance.
Published in:	Proceedings of the 12th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1999) September 14 - 17, 1999 Nashville, TN
Pages:	2027 - 2032
Cite this article:	I.Fossen, Bjørnar Vikand Thor, "A Nonlinear Observer for Integration of GPS and INS Attitude," Proceedings of the 12th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1999), Nashville, TN, September 1999, pp. 2027-2032.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In