This paper presents the problematic of localization under the IMM (Interacting Multiple Model) approach. The localization is often tackled under the accuracy problem. In order to achieve the goal of the assessment of an accurate ego localization, one often uses the merging of data coming from both exteroceptive and proprioceptive sensors. In our approach, we don't focuss much on the accuracy, but more especially on both the confidence and the robustness of this positioning. In fact, the IMM approach is based on the discretization of the vehicle evolution space into simple maneuvers, represented each by a simple dynamic model. In order to reach our objective using IMM, we assume that at every time period the true mode under which the vehicle goes is represented. After a review of different traditional filters used in vehicle localization, an IMM method is proposed and the comparison is mainly based on some robustness criteria that are presented in this paper.
A Multiple Model Localization System for Outdoor Vehicles
2007 IEEE Intelligent Vehicles Symposium ; 1050-1055
2007-06-01
777573 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
A Multiple Model Localization System for Outdoor Vehicles
British Library Conference Proceedings | 2007
|An Outdoor Localization System Based on SigFox
IEEE | 2018
|LANE SELF-LOCALIZATION SYSTEM USING MULTIPLE CAMERAS FOR AUTONOMOUS DRIVING VEHICLES
Europäisches Patentamt | 2021
|LANE SELF-LOCALIZATION SYSTEM USING MULTIPLE CAMERAS FOR AUTONOMOUS DRIVING VEHICLES
Europäisches Patentamt | 2019
|