Multiple model framework of adaptive extended kalman filtering for predicting vehicle location
C. Barrios, Henry Himberg, Yuichi Motai, Ahmed K. Sadek · 2006
A multiple-model framework of adaptive extended Kalman filters (EKF) for predicting vehicle position with the aid of Global Positioning System (GPS) data is proposed to improve existing collision avoidance systems. A better prediction model for vehicle positions provides more accurate collision warnings in situations that current systems can not handle correctly. The multiple model adaptive estimation system (MMAE) algorithm is applied to the integration of GPS measurements to improve the efficiency and performance. This paper evaluates the multiple-model system in different scenarios and compares it to other systems before discussing possible improvements by combining it with other systems for predicting vehicle location