Experimental validation of attitude and rate-sensor bias filter using range-difference measurements. (April 2018)

• Record Type: Journal Article
• Authors: Jørgensen, Erlend K.; Fossen, Thor I.; Schjølberg, Ingrid; Esperança, Paulo T.T.
• Is Part Of: Control engineering practice. Volume 73(2018)
• Page Start: 112
• Page End: 123
• Publication Date: 2018-04
• Subjects: Attitude estimation -- Rate-sensor bias estimation -- Non-linear filtering -- Exogenous Kalman Filter -- Linearized Kalman Filter; Automatic control -- Periodicals; 629.89
• DOI: 10.1016/j.conengprac.2018.01.002
• Languages: English
• ISSNs: 0967-0661

Abstract

Abstract: This paper considers the problem of constructing a filter for estimating attitude and rate-sensor bias, that has both proven stability and close-to-optimal performance with respect to noise. The filter is based on measuring the difference in time of arrival for signals sent from three or more known, fixed positions to two or more receivers on the vehicle. An inertial measurement unit is also used, both rate-sensor and accelerometer measurements, and a position estimate is needed, generated from depth and time of arrival measurements. The vectors between receivers on the vehicle are assumed to be known in the body frame, and are calculated in the inertial frame through an algebraic transformation. These vectors are used as input for a non-linear observer along with rate-sensor and accelerometer data, estimating Euler angles and rate-sensor bias. These estimates are used as a linearization point for a Linearized Kalman Filter, taking the full non-linear system into account. Two experiments are run, and the filter is compared to an Extended Kalman Filter, and a non-implementable Linearized Kalman Filter using the true state as linearization point. Highlights: An experimental validation of a filter estimating attitude and rate-sensor bias is provided. The filter is based on the eXogenous Kalman Filter principle. The computational burden of the suggested filter is about twice that of the Extended Kalman Filter. The filter is compared in experiments to an Extended Kalman Filter. Two experiments are conducted. The filter converges quickly and the stationary error is small.