Détail de l'auteur
Auteur B. Gundlich |
Documents disponibles écrits par cet auteur (1)
Ajouter le résultat dans votre panier Affiner la recherche Interroger des sources externes
Statistische Untersuchung ganzzahliger und reellwertiger unbekannter Parameter im GPS-Modell / B. Gundlich (2002)
Titre : Statistische Untersuchung ganzzahliger und reellwertiger unbekannter Parameter im GPS-Modell Titre original : [Statistical inference on integer and real valued unknown parameters in the GPS model] Type de document : Thèse/HDR Auteurs : B. Gundlich, Auteur Editeur : Munich : Bayerische Akademie der Wissenschaften Année de publication : 2002 Collection : DGK - C Sous-collection : Dissertationen num. 549 Importance : 65 p. Format : 21 x 30 cm ISBN/ISSN/EAN : 978-3-7696-9588-5 Note générale : Bibliographie Langues : Allemand (ger) Descripteur : [Vedettes matières IGN] Géodésie spatiale
[Termes IGN] ambiguïté entière
[Termes IGN] axiome de Bayes
[Termes IGN] données GPS
[Termes IGN] inférence
[Termes IGN] mesurage de phase
[Termes IGN] mesurage de pseudo-distance
[Termes IGN] méthode de Monte-Carlo
[Termes IGN] résolution d'ambiguïté
[Termes IGN] traitement de données GNSSIndex. décimale : 30.61 Systèmes de Positionnement par Satellites du GNSS Résumé : (Auteur) This thesis investigates statistical inference in a model with real valued and integer unknown parameters. This model is given for example in case of GPS (Global Positioning System) which provides pseudorange measurements and the observation of carrier phases. When observing carrier phases the integer numbers of full cycles are unknown. These integer ambiguities have to be considered in a statistical inference. Usually the float solution and the fixed solution are applied. Using the float solution the integer ambiguities are considered as real valued unknown parameters. In the fixed solution they are completely or partly estimated as integers and then fixed. The remaining parameters are estimated given the integer ambiguities. Therefore the fixed solution has to be validated. The ambiguity acceptance test investigates, if the model with integer ambiguities might be accepted. The integer estimation of the ambiguities is compared with other possible integer ambiguities in a discrimination test. In this thesis the GPS model is consistently considered as a model containing integer and real valued unknown parameters in order to avoid the extreme float and fixed solution. Applying traditional statistics and Bayesian statistics the unknown parameters are estimated, hypotheses are tested and confidence regions computed. This leads to new validation procedures concerning the ambiguity acceptance test and the discrimination test. Estimating confidence regions is computationally demanding, therefore an approximation is developed by Bayesian statistics. Bayesian statistics offers the possibility to use prior information, which is used to express the knowledge about the integer ambiguities in a model with only real valued parameters. By this means solutions are developed which only in extreme cases lead to the float solution or to the solution in the model with integer and real valued parameters. An example with real GPS data is given. Numéro de notice : 13099 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Thèse étrangère Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=54882 Réservation
Réserver ce documentExemplaires(2)
Code-barres Cote Support Localisation Section Disponibilité 13099-02 30.61 Livre Centre de documentation Géodésie Disponible 13099-01 30.61 Livre Centre de documentation Géodésie Disponible