Institutional Repository
Technical University of Crete
Technical University of Crete
EN
|
EL
| URI: | http://purl.tuc.gr/dl/dias/C32F4F8F-5490-4620-B94D-721257EF8A9A | ||
| Year | 2025 | ||
| Type of Item | Diploma Work | ||
| License |
|
||
| Bibliographic Citation | Christos Pavlopoulos, "Modeling and prediction of photovoltaic system performance using digital twin technology", Diploma Work, School of Electrical and Computer Engineering, Technical University of Crete, Chania, Greece, 2025 https://doi.org/10.26233/heallink.tuc.102277 | ||
| Appears in Collections |
Solar energy is emerging as the simplest of the forms of Renewable Energy Sources (RES) and photovoltaic (PV) systems are expected to become the main form of production in countries with abundant solar irradiation potential, such as Greece. The simplicity of these systems is due to the easy prediction of environmental conditions during the year, high reliability, and minimal maintenance needs of the equipment. This simplicity is the source of inspiration for this thesis. Using the emerging technology of digital twins, i.e., the creation of a digital clone to monitor and control a real system, the focus of the thesis is to create a simple, easily adaptable model that can serve as base for other models. This paper shows the analysis of electricity generation and weather data from an operating photovoltaic park, the process of training the Autoregressive Moving Average Model with Exogenous Variables (ARMAX) using machine learning methods such as decision trees, and its simulation in the MATLAB environment. A detailed comparison of the model predictions is presented for data from different time periods. The analysis includes both data used for training the model and data not previously introduced to it. The predictions are compared with actual measurements to evaluate the model accuracy and effectiveness. Finally, future extensions and usage of the model are discussed.
Copyright © DIAS 2013
