Skip to Main content Skip to Navigation

Apports des techniques de filtrage non linéaire pour la navigation avec les systèmes de navigation inertielle et le GPS

Abstract : Navigation science has long been the prerogative of the army but mass-market locationbased services are now part of everyday life. The design of reliable and robust positioning systems is of the utmost importance especially for safety critical applications (for instance civil aviation). This thesis focuses on GPS and Inertial Navigation Systems (INS) and investigates two strands of research to enhance navigation performance. On the one hand, we have studied hybridization techniques, which take advantage of the complementarity of different navigationsystems. On the other hand, this work is also dedicated to developing robust estimation algorithms to recover a mobile trajectory from a set of sensor measurements. Particle filters are an interesting alternative to the well-known Kalman filter since they can be applied to non linear and possibly non Gaussian state space models. They consist in sequentially propagating samples that represent a possible value of the state vector at each time instant. In this work, we have considered using Particle Filters to solve the non linear problems of GPS/INS tight and very tight couplings. Thus, we have proposed an extension to the well-known Regularised Particle filter, which exhibits good performance both in terms of robustness and accuracy. Finally, a second contribution of this work is the development of Particle Filtering algorithms to track a mobile trajectory while jointly detecting/estimating potential degradations on the GPS measurements.
Keywords : GPS/INS
Complete list of metadatas

Cited literature [88 references]  Display  Hide  Download
Contributor : Laurence Porte <>
Submitted on : Thursday, July 2, 2020 - 10:56:25 PM
Last modification on : Friday, July 10, 2020 - 2:28:35 PM
Long-term archiving on: : Thursday, September 24, 2020 - 6:27:14 AM


Files produced by the author(s)


  • HAL Id : tel-02888313, version 1



Audrey Giremus. Apports des techniques de filtrage non linéaire pour la navigation avec les systèmes de navigation inertielle et le GPS. Traitement du signal et de l'image [eess.SP]. ENSAE, 2005. Français. ⟨tel-02888313⟩



Record views


Files downloads