Fuzzy adaptive Kalman filtering for DR/GPS
Santong Zhang, Xueye Wei · 2004
In this paper, a novel method of multi-sensor data fusion based on the Adaptive Fuzzy Kalman Filter is presented. This method is applied in fusing position and orientation (or direction) signals from Dead Reckoning (DR) system and the Global Positioning System (GPS) for landing vehicle navigation. The Extended Kalman Filter (EKF) and the characteristics of the measurement noise are modified by using the Fuzzy Adaptive system, and Fuzzy Adaptive system is based on a covariance matching technique. It is compared with the performance of a regular EKF. It is demonstrated that Fuzzy Adaptive Kalman Filter is better (more accurate) than the EKF, and the algorithm is not complex. It is important to improve the accuracy of the vehicle navigation system.