User: Guest  Login
Author(s):
Kiral, Alperen
Title:
Automated Calibration Methods for Digital Production Twins
Abstract:
Digital twin can simply be defined as a virtual replica of a physical system based on acquired online data. It allows companies to monitor physical assets through the entire lifecycle, thereby creating a new business strategy that decreases costs and increases quality & customer satisfaction. But be that as it may, if a digital twin is not relatively accurate, it can do harm, especially during operation. In the manufacturing industry, an operator, who usually is not a simulation expert, needs to update digital twins not only for an initial calibration but also as a regular maintenance. By minimizing manual efforts in these processes, cost of many labor hours and possible inaccuracies can be correspondingly reduced and prevented. To overcome this challenge in a generalized way, a calibration software is being developed in this thesis which integrates field data with a computer vision-aided camera system to automatically calibrate specifically digital production twins. The calibration methods are applied to a FischerTechnik model plant coupled with real automation hardware and software. Many challenging experiments are conducted to test the capabilities of the code, by which the performance of the code is measured. Overall, the methods calibrate the digital twins accurately and efficiently.
Keywords:
LOCenter;
Subject:
ALL Allgemeines
Supervisor:
Braun, A. and Borrmann, A. and Wehr, J.
Year:
2021
Year / month:
2021-04
Month:
Apr
University:
Technische Universität München
 BibTeX