Short-term prediction of electrical energy consumtion using regression models
| Autor: | Kuprešak, Tea |
|---|---|
| Přispěvatelé: | Matuško, Jadranko |
| Jazyk: | chorvatština |
| Rok vydání: | 2021 |
| Předmět: |
Artificial intelligence
regression model Regression Learner App TEHNIČKE ZNANOSTI. Elektrotehnika Kratkoročna predikcija potrošnje PSI Control Livno Branch Poslovnica Livno TECHNICAL SCIENCES. Electrical Engineering Short-term consumption prediction umjetna inteligencija statistical methods statističke metode regresijski model Matlab |
| Popis: | Većina metoda predikcije potrošnje električne energije koristi statističke metode ili algoritme s umjetnom inteligencijom poput regresijskih postupaka, neuronskih mreža, neizrazite logike i ekspertnih sustava. U radu je analizirana primjena regresijskih postupaka za kratkoročnu predikciju potrošnje električne energije u Poslovnici Livno, korištenjem dva različita programska okruženja. Temeljem dobivenih podataka o potrošnji električne energije, tipu dana i blagdana, te meteoroloških podataka koji služe kao prediktori za prognozu potrošnje, dobiven je regresijski model unutar programskog okruženja PSI Control. Model prognoze ponaša se kao prilagodba putem Kalmanova filtra. Regresijski model implementiran je i unutar programskog okruženja Matlab te je obavljena statistička i grafička analiza predikcije potrošnje odabranog regresijskog modela. Most methods of predicting electricity consumption use statistical methods or algorithms with artificial intelligence such as regression procedures, neural networks, fuzzy logic, and expert systems. In this thesis was analyzed use of regression procedures for short-term prediction of electricity consumption in the Livno Branch, using two different software environments. Based on the obtained data of electricity consumption, type of days and holidays, and meteorological data that serve as predictors for the consumption forecast, a regression model was obtained within the PSI Control software environment. The forecast model acts as an adjustment via the Kalman filter. The regression model was also implemented within the Matlab software environment, and a statistical and graphical analysis of the consumption prediction of the selected regression model was performed. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|