Self-localization for a mobile vehicle by sensor fusion

Comparison of structure between grid point observer and extended Kalman filter

Abstract

An Extended Kalman filter (EKF) is a major sensor fusion technique for the self-localization of mobile vehicles. To calculate Kalman gains, variance matrices are necessary. However, it is laborious to obtain variance matrices. To solve this problem, we introduce a grid point observer and compare the structure of grid point observer with that of EKF.