NORMA eResearch @NCI Library

Understanding the Impact of COVID-19 on Electrical Demand

Nichanian, Stephane (2020) Understanding the Impact of COVID-19 on Electrical Demand. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
Preview
PDF (Master of Science)
Download (3MB) | Preview
[thumbnail of Configuration manual]
Preview
PDF (Configuration manual)
Download (1MB) | Preview

Abstract

The COVID-19 pandemic has fundamentally changed our society’s behaviour. Along with these changes, electrical consumption has also been impacted in ways never expected before. This research has highlighted the main changes in electrical consumption statistics during lockdown: the general decrease of up to 20% electrical demand, the assimilation of weekend and weekdays and the shift of daily activities towards later hours of the day. From a technical standpoint this study has explored two powerful regression techniques, Generalized Additive Models and ARIMA polynomial regression. The choice of regression techniques is justified by the control it gives over predictor variables and the possibility for manual tuning of model parameters. An important choice of forecasting methodology was to create a separate model for each hour of the day, thus creating accurate representation of the complex intra-daily seasonality. Additionally, the introduction of lagged demand variables was instrumental in reducing autocorrelation as well as increasing model accuracy. The study has found GAM models to perform slightly better on short-term forecast (24 days ahead) with a total MAPE of 2.24 over the 5 dates considered against 2.92 for ARIMA regression. A mid-term forecast was also implemented for a period of 3 months where the GAM model significantly outperformed the ARIMA regression with an MAPE of 2.31 against 4.08 for ARIMA regression. We do however note, significantly longer processing times for GAM models due to complex smoothing functions. We also note that there is no substantial degradation of prediction accuracy from a one day ahead forecast to a 3-month forecast, which validates the usage of GAM models for the lockdown period simulation.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
Q Science > QA Mathematics > Computer software
T Technology > T Technology (General) > Information Technology > Computer software
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Dan English
Date Deposited: 22 Jan 2021 15:52
Last Modified: 22 Jan 2021 15:52
URI: https://norma.ncirl.ie/id/eprint/4458

Actions (login required)

View Item View Item