Modified Extended Kalman Filter Technique for Car Navigation in Urban Environment with Limited GPS Visibility

GPS 위성의 가시성이 제한을 받는 도심지 환경하에서의 차량항법을 위한 변형된 확장칼만필터기법

In this paper, Modified GPS Kalman filter algorithms which allow user to estimate its position when the number of visible GPS satellites becomes less than four are presented. They are derived using the previous estimation of altitude and clock bias. Thus, it is possible to estimate 3-dimensional user position even when only two GPS satellites are visible. The algorithms are ideally suited to car navigation in urban areas where lack of GPS visibility is the major problem because of the frequent blockage of the GPS signals by tall buildings and other structures. Simulation results in this paper show that modified GPS Kalman filter provide better performances than a general GPS Kalman filter or any other instantaneous GPS solution algorithm, especially in the case which the number of visible GPS satellites becomes less than four.