Interacting multiple model for improving the precision of vehicle-mounted global position system

摘要

To improve the precision of low-cost, Vehicle-mounted global position system (GPS), this paper presents the multi-source information fusion algorithm of vehicle navigation, which is based on the interacting multiple model (IMM). Considering vehicle kinematics and dynamic characteristics, as well as its braking capacity in extreme accelerating situations, this study establishes the clothoid model, which shows longitudinal and lateral constant speed motion, and the Adaptive Current Statistics (ACS) model, which shows the acceleration of a vehicle. Through the parametric estimation of interactivity, filtering, and updating of probability, the vehicle trajectory is predicted within a period of time, and high-precision dead-reckoning is therefore achieved. Comparative analysis shows that the above algorithm can improve the precision of vehicle-mounted GPS/inertial navigation system. (C) 2015 Elsevier Ltd. All rights reserved.