Impact of coronavirus disease 2019 on electricity demand and the unit commitment problem: a long–short-term memory-based machine learning approach. Issue 11 (2nd November 2022)

• Record Type: Journal Article
• Authors: Elkamel, Marwen; Ahmadian, Ali; Zheng, Qipeng P.
• Is Part Of: Engineering optimization. Volume 54:Issue 11(2022)
• Page Start: 1835
• Page End: 1852
• Publication Date: 2022-11-02
• Subjects: COVID-19 -- unit commitment -- electricity demand -- machine learning -- optimization; Engineering design -- Periodicals; Mathematical optimization -- Periodicals; 620.0042
• DOI: 10.1080/0305215X.2021.1961762
• Languages: English
• ISSNs: 0305-215X

Abstract

Abstract : Coronavirus disease 2019 (COVID-19) has affected many behaviours and aspects of society. Electricity consumption has been considerably affected by the pandemic, with significant effects on the electricity load demand profile. In this article, the impact of COVID-19 on electricity demand in the state of Florida is investigated through a novel machine learning technique. The LSTM technique shows good accuracy in forecasting the load profiles for all days studied (weekdays and weekends) and also before and during the pandemic. The UC problem is solved considering the load profiles, and the impact of COVID-19 on power plant scheduling is evaluated. The simulation results show an increase in residential demand for electricity at weekends, while both residential and commercial demand are reduced during weekdays. Therefore, the operating cost of a weekday in 2020 was lower than that in 2019, while the operating cost of a weekend was higher in 2020 than in 2019.