Datenbestand vom 23. März 2024

Warenkorb Datenschutzhinweis Dissertationsdruck Dissertationsverlag Institutsreihen     Preisrechner

aktualisiert am 23. März 2024

ISBN 9783843914161

84,00 € inkl. MwSt, zzgl. Versand


978-3-8439-1416-1, Reihe Ingenieurwissenschaften

Mohammed Khider
Multisensor-Based Positioning for Pedestrian Navigation

159 Seiten, Dissertation Universität Ulm (2013), Softcover, A5

Zusammenfassung / Abstract

A rapidly growing market for pedestrian location-based services has developed in recent years. Offering the pedestrian the right service, at the right time and in the right place requires accurate knowledge of their position. Global navigation satellite systems (GNSSs) - the best known type of positioning system - fail to provide accurate positioning in indoor and urban canyon environments due to multipath propagation and signal blockage. A substantial quantity of work has recently been carried out in developing positioning approaches that are reliable in all environments. As all single-sensor positioning systems fail, multisensor positioning - where information from two or more positioning sources is combined - represents the state-of-the-art solution. Bayesian positioning algorithms have shown promising results in optimally combining information from different positioning sources.

The goal of this work is the development of an optimal pedestrian position estimator able to provide sufficient accuracy and availability in both indoor and outdoor environments. To this end, the use of GNSSs in multisensor positioning approaches has been enhanced through appropriately combining satellite-to-user range measurements with human odometry and position information from other sources. Using satellite-to-user range measurements instead of GNSS receiver position solutions reduces the number of satellite signals required. Moreover, it allows the incorporation of range measurement error models. With the aim of developing an optimal position estimator, two novel pedestrian movement models able to realistically represent the stochastic nature of pedestrian movement have been developed. Incorporating such movement models into Bayesian position estimators is beneficial as they allow pedestrian position and direction in the event of measurement unavailability to be predicted, and moreover help filter erroneous sensor outputs. An optimal Bayesian position estimator has been developed incorporating state-of-the-art fusion algorithms, the movement models developed, appropriately modeled satellite-to-user range measurements, human odometries and other position-related measurements.