Design and implementation of an adaptive extended Kalman filter with interval type-3 fuzzy set for an attitude and heading reference system

Abstract

There are various techniques for estimating dynamic systems. Two of the most commonly used filters are the linear Kalman filter and the extended Kalman filter. However, these methods have some limitations since both the dynamic equations and the measurement equations need to be linearized. In addition, uncertainties in the dynamic equations and noise in the measurement equations must be considered when implementing these algorithms, which can be seen in the accurate determination of the covariance matrix of the process and measurement noise. This study focuses on improving the accuracy of the attitude and heading reference system through improved extended Kalman filter performance. This improvement is achieved by adjusting the covariance matrix with respect to the measurement noise using the interval type-3 fuzzy set in conjunction with the covariance matching technique. The novel algorithm is tested on a ground vehicle in an urban environment and its effectiveness compared to the traditional extended Kalman filter is demonstrated.