Bayesian navigation filters are broadly exploited in precise state estimation for kinematic applications such as vehicular positioning and navigation. Among these, Particle Filter (PF) has been shown as a valuable solution to support hybrid positioning algorithms such as sensor fusion to Global Navigation Satellite System (GNSS) and Cooperative Positioning (CP). Despite of an increased computational complexity w.r.t. conventional Kalman Filters (KFs), an effective weighting of the input measurements generally provides an improved accuracy of the output estimate. In the framework of the Differential GNSS (DGNSS) CP, this work presents an algorithm for the automated selection of the most appropriate error models for the tight-integration of non-stationary Differential GNSS (DGNSS) collaborative inter-agent distances. A model switching technique named Automated Adaptive Likelihood Switch (AALS) is proposed for a Cognitive Particle Filter (C-PF) architecture, based on the real-time approximation of the statistics of the inter-agent distances errors. The results achieved through realistic simulations demonstrated the effectiveness of the proposed solution in terms of error model selection. Therefore, an improvement of the position estimation accuracy was observed, since the cases in which DGNSS-CP would degrade performance due to possible mismodelling of the selected likelihood function are avoided.
Adaptive Bayesian State Estimation Integrating Non-stationary DGNSS Inter-Agent Distances / Zocca, Simone; Minetto, Alex; Dovis, Fabio. - ELETTRONICO. - (2021). (Intervento presentato al convegno 2021 IEEE 93rd Vehicular Technology Conference (VTC2021-Spring) tenutosi a Tampere, Finland nel 25-28 April) [10.1109/VTC2021-Spring51267.2021.9448952].
Adaptive Bayesian State Estimation Integrating Non-stationary DGNSS Inter-Agent Distances
Simone Zocca;Alex Minetto;Fabio Dovis
2021
Abstract
Bayesian navigation filters are broadly exploited in precise state estimation for kinematic applications such as vehicular positioning and navigation. Among these, Particle Filter (PF) has been shown as a valuable solution to support hybrid positioning algorithms such as sensor fusion to Global Navigation Satellite System (GNSS) and Cooperative Positioning (CP). Despite of an increased computational complexity w.r.t. conventional Kalman Filters (KFs), an effective weighting of the input measurements generally provides an improved accuracy of the output estimate. In the framework of the Differential GNSS (DGNSS) CP, this work presents an algorithm for the automated selection of the most appropriate error models for the tight-integration of non-stationary Differential GNSS (DGNSS) collaborative inter-agent distances. A model switching technique named Automated Adaptive Likelihood Switch (AALS) is proposed for a Cognitive Particle Filter (C-PF) architecture, based on the real-time approximation of the statistics of the inter-agent distances errors. The results achieved through realistic simulations demonstrated the effectiveness of the proposed solution in terms of error model selection. Therefore, an improvement of the position estimation accuracy was observed, since the cases in which DGNSS-CP would degrade performance due to possible mismodelling of the selected likelihood function are avoided.File | Dimensione | Formato | |
---|---|---|---|
09448952.pdf
non disponibili
Descrizione: Articolo disponibile su IEEExplore
Tipologia:
2a Post-print versione editoriale / Version of Record
Licenza:
Non Pubblico - Accesso privato/ristretto
Dimensione
2.55 MB
Formato
Adobe PDF
|
2.55 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
VTC21-Spring-Manuscript.pdf
accesso aperto
Descrizione: Articolo disponibile su IEEExplore
Tipologia:
2. Post-print / Author's Accepted Manuscript
Licenza:
PUBBLICO - Tutti i diritti riservati
Dimensione
719.03 kB
Formato
Adobe PDF
|
719.03 kB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/11583/2907872