University of Twente Student Theses


Context Aware GPS Error Correction

Ong-A-Fat, K.K. (2017) Context Aware GPS Error Correction.

[img] PDF
Abstract:One of the most significant errors encountered in GPS data are signal multipath errors. The classification and correction of these type of errors are considered in this research. The aim of this thesis is to describe a framework that utilizes machine learning techniques in combination with the characteristics of signal multipath errors to automatically classify and correct these errors. The framework is designed to classify signal multipath errors in data from various fields of expertise by using a semi-supervised machine learning approach that uses the unknown dataset to train its classifier. By doing so the framework is applicable on a vast range of different datasets. The framework is validated using a specific case study with data from asphalt paving projects. These projects contain the GPS trajectories of various rollers that were used during the paving process. The framework performed well on classifying and correcting regular signal multipath errors but had difficulty to identify signal multipath errors that had irregular patterns. The classification accuracy on correctly classifying signal multipath errors on the created testing set was 92%. The framework showed its capabilities of automated signal multipath error classification and correction, however future research is needed to create a practically applicable solution.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Computer Science MSc (60300)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page