An adaptive strong tracking Kalman filter for position and orientation system

A position and orientation systems (POS) plays an important role in aerial mapping applications. It integrates the inertial navigation system and global positioning system to provide high-precision position, velocity and attitude for various aerial mapping sensors. However, in severe environment of temperature, magnetic field and vibration in the application of aerial mapping, the precision of gyroscopes and accelerometers may degrade. The traditional Kalman filter may perform poorly when the model of gyroscope and accelerometer errors is uncertain. This paper highlights the use of multiple fading factors for a strong tracking Kalman filter (STKF) to accommodate the model uncertainty of gyroscope and accelerometer errors. Through utilizing the information of the sensitivity matrix of a two-stage Kalman filter, the multiple fading factors are obtained adaptively. Therefore, a more accurate covariance matrix is obtained in the proposed algorithm, and a better state tracking ability is achieved than with the Kalman filter and the STKF. Finally, a flight experiment is demonstrated to validate the effectiveness of the proposed algorithm. It is shown from the experimental results that the proposed algorithm can more accurately estimate the time-varying errors of gyroscopes and accelerometers than Kalman filters or the STKF; the accuracy of position, velocity and attitude of POS is also improved correspondingly.