Autor: Sebastian RICHTER
Informatyka, semestr 3
specjalność: Oprogramowanie Systemowe
Wydział Automatyki, Elektroniki i Informatyki
Politechnika Śląska
Promotor: dr hab. inż. Adam DOMAŃSKI, prof. PŚ
Predykcja produkcji energii elektrycznej z paneli fotowoltaicznych na podstawie prognozy pogody.
Celem projektu jest stworzenie programu, który będzie przewidywać ilość wyprodukowanej energii elektrycznej przez panele fotowoltaiczne na podstawie prognozy pogody dla lokalizacji, w których zostały one zamontowane. Prognozę produkcji energii elektrycznej będzie można przeprowadzić w kliku różnych perspektywach czasowych: ultrakrótkoterminowej (od kilku minut do kilku godzin naprzód), krótkoterminowej (do kilkudziesięciu godzin naprzód) oraz długoterminowej (od kilku do kilkudziesięciu dni naprzód). Algorytm predykcji produkcji energii z paneli fotowoltaicznych zostanie zaimplementowany przy użyciu metod sztucznej inteligencji w taki sposób, by błędy w uzyskanych wynikach były jak najmniejsze. Wykresy będące rezultatem badań będą obrazować ilość wyprodukowanej energii elektrycznej w zależności od różnych parametrów opisujących pogodę (np. temperatura powietrza, wilgotność, ciśnienie atmosferyczne).
- Python 3.11.1
- TensorFlow (Keras)
- IDE: Visual Studio Code
- system kontroli wersji Git
Repozytorium projektu zostało udostępnione na Licencji MIT.
Author: Sebastian RICHTER
Informatics, 3rd semester
specialization: System Software
Faculty of Automatic Control, Electronics and Computer Science
Silesian University of Technology
Promoter: dr hab. inż. Adam DOMAŃSKI, prof. PŚ
Prediction of electricity generation from photovoltaic modules based on weather forecast.
The aim of the project is creating a program that will predict the amount of electricity produced by photovoltaic panels based on weather forecasts for the locations where they were installed. It will be possible to forecast electricity generation in several different time frames: the ultra-short-term (from a few minutes to a few hours ahead), the short-term (up to a few dozen hours ahead) and the long-term (from a few to several dozen days ahead). The algorithm for predicting the production of energy from photovoltaic panels will be implemented using artificial intelligence methods in such a way that errors in the obtained results are as small as possible. The graphs resulting from the tests will show the amount of electricity produced depending on different parameters describing the weather (e.g. air temperature, humidity, atmospheric pressure).
- Python 3.11.1
- TensorFlow (Keras)
- IDE: Visual Studio Code
- Git version control system
Licensed under the MIT License.