https://doi.org/10.1051/epjpv/2024023
Review
Benchmarking photovoltaic plant performance: a machine learning model using multi-dimensional neighbouring plants
1
Green Power Monitor a DNV company, Gran Via de les Corts Catalanes 130, Barcelona, Spain
2
DNV Denmark, Tuborg Parkvej 8, Hellerup, Denmark
* e-mail: gaetana.anamiati@dnv.com
Received:
30
November
2023
Accepted:
4
July
2024
Published online: 13 August 2024
The goal of this study is to monitor the performance of a photovoltaic plant by comparing its power output against others with similar characteristics, referred to as neighbours. The purpose of the 15 neighbours is to have the best reference for the performance of the plant in question, in other words, how the plant under analysis performs compared to the 15 neighbours. A machine learning model based on a feed forward Neural Network was employed to model power production as a function of environmental signals and the sun's position. Here, the data from the neighbours are used to train the model and the data from the plant under analysis are used to evaluate the model and predict the power output. Once the power is predicted, the performance ratio of the plant is calculated. The procedure has been tested and validated at several plants for three different cases and the numerical results highlight how the model is able to identify under/over performing plants. Therefore the developed strategy provides industries a valid tool on the correct functioning of a plant.
Key words: Solar energy / photovoltaic plants / performance / machine learning
© G. Anamiati et al., Published by EDP Sciences, 2024
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
