A new robust cooperative-reactive Filter for Vehicle Localization: The Extended Kalman Particle Swarm 'EKPS'

Abstract : This paper introduces a proposal for a collaborative intelligent localization algorithm inspired from the Particle Swarm Optimization (PSO) technique and applied to highly dynamic road vehicle localization. This approach performs a reactive cooperative vehicle localization by considering a PSO of the vehicle position in a dynamic environment with an adaptive dynamic 'fitness' function. In order to manage the uncertainties, the PSO algorithm is coupled with an Extended Kalman Filter (EKF). This new localization approach is tested and validated using real world data obtained from embedded sensors (GPS, INS, Odometer, Gyrometer, Steering wheel angle sensor and a Centimetrik RTK GPS) in comparison with the classical EKF performances. The first results obtained are better in terms of accuracy and robustness.
Document type :
Conference papers
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-00857236
Contributor : Ifsttar Cadic <>
Submitted on : Tuesday, September 3, 2013 - 11:05:13 AM
Last modification on : Monday, February 10, 2020 - 11:42:10 AM

Identifiers

Collections

Citation

Adda Redouane Ahmed Bacha, Dominique Gruyer, Said Mammar. A new robust cooperative-reactive Filter for Vehicle Localization: The Extended Kalman Particle Swarm 'EKPS'. IEEE Intelligent Vehicles Symposium (IV 2013), Jun 2013, Gold Coast, Australia. pp.195--200, ⟨10.1109/IVS.2013.6629470⟩. ⟨hal-00857236⟩

Share

Metrics

Record views

202