Advanced signal processing using automated uncertainty propagation - An educational approach
Year: 2024
Type: article
Abstract: Automated uncertainty propagation is a powerful tool that allows to focus on the interpretation and analysis of the uncertainty rather than on its calculation. A teaching concept based on the use of a software toolbox for automated uncertainty propagation have been previously developed. This work introduces Bayesian Linear Minimum Mean Square Estimation with automated uncertainty propagation, proposing an extension of the teaching concept, towards the use of uncertainty in sensor and data fusion. The toolbox provides the methods linearization, Unscented Transform and Monte Carlo for uncertainty propagation. Since the code for all three variants is essentially equivalent and compact, concepts as the Extended Kalman Filter and the Unscented Kalman Filter are easily implemented following the same approach as for the already introduced uncertainty quantification. The proposed approach simplifies the integration of advanced topics such as late fusion or posterior estimation of previous states in measurement science education.
Source: Measurement Sensors
Authors: Hubert Zangl, Dailys Arronde Pérez
Institutions: Silicon Austria Labs (Austria), University of Klagenfurt
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $1,800
OpenAlex ID: W4405863934