eduzhai > Applied Sciences > Engineering >

Using Mobility for Electrical Load Forecasting During the COVID-19 Pandemic

  • Save

... pages left unread,continue reading

Document pages: 8 pages

Abstract: The novel coronavirus (COVID-19) pandemic has posed unprecedented challengesfor the utilities and grid operators around the world. In this work, we focuson the problem of load forecasting. With strict social distancing restrictions,power consumption profiles around the world have shifted both in magnitude anddaily patterns. These changes have caused significant difficulties inshort-term load forecasting. Typically algorithms use weather, timinginformation and previous consumption levels as input variables, yet they cannotcapture large and sudden changes in socioeconomic behavior during the pandemic.In this paper, we introduce mobility as a measure of economic activities tocomplement existing building blocks of forecasting algorithms. Mobility dataacts as good proxies for the population-level behaviors during theimplementation and subsequent easing of social distancing measures. The majorchallenge with such dataset is that only limited mobility records areassociated with the recent pandemic. To overcome this small data problem, wedesign a transfer learning scheme that enables knowledge transfer betweenseveral different geographical regions. This architecture leverages thediversity across these regions and the resulting aggregated model can boost thealgorithm performance in each region s day-ahead forecast. Through simulationsfor regions in the US and Europe, we show our proposed algorithm can outperformconventional forecasting methods by more than three-folds. In addition, wedemonstrate how the proposed model can be used to project how electricityconsumption would recover based on different mobility scenarios.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...