This work investigates the use of predicting algorithms for energy consumption in the tertiary sector. Tertiary sector is fast-growing, in fact it used 23% of Italian electricity in 2000 reaching the 35% in 2016. The focus of this paper is on banker institutions spread across the Italian country. Several algorithms are taken into account and compared in order to find the best solution. The proposed algorithms underwent a training period where the parameter with higher impact on the overall consumption are taken into account. Once the model was trained, the last year of historical data was used to verify the quality of the proposed approach. Final remarks discuss possible algorithm refinement as well as its use for the quick detection and correction of anomalies in the energy use profile curves.
Predictive algorithms for energy performance evaluation of banking institutions / Fondriest, M.; Macchitelli, G.; Stancari, S.; Montanari, D.; Fiorini, C.; Anceschi, G.; Pedrazzi, S.; Allesina, G.. - 2191:(2019), p. 020077. (Intervento presentato al convegno 74th Conference of the Italian Thermal Machines Engineering Association, ATI 2019 tenutosi a Department of Engineering "Enzo Ferrari" of the University of Modena and Reggio Emilia, ita nel 2019) [10.1063/1.5138810].
Predictive algorithms for energy performance evaluation of banking institutions
Pedrazzi S.;Allesina G.
2019
Abstract
This work investigates the use of predicting algorithms for energy consumption in the tertiary sector. Tertiary sector is fast-growing, in fact it used 23% of Italian electricity in 2000 reaching the 35% in 2016. The focus of this paper is on banker institutions spread across the Italian country. Several algorithms are taken into account and compared in order to find the best solution. The proposed algorithms underwent a training period where the parameter with higher impact on the overall consumption are taken into account. Once the model was trained, the last year of historical data was used to verify the quality of the proposed approach. Final remarks discuss possible algorithm refinement as well as its use for the quick detection and correction of anomalies in the energy use profile curves.
Pubblicazioni consigliate
I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris
