eduzhai > Applied Sciences > Engineering >

Deep Learning Based Load Forecasting from Research to Deployment -- Opportunities and Challenges

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 6 pages

Abstract: Electricity load forecasting for buildings and campuses is becomingincreasingly important as the penetration of distributed energy resourcesgrows. Efficient operation and dispatch of DERs requires reasonably accurateprediction of future energy consumption in order to conduct near-real-timeoptimized dispatch of on-site generation and storage assets. Load forecastinghas traditionally been done by electric utilities for load pockets spanninglarge geographic areas and therefore has not been a common practice in thebuildings and campuses operational arena. Given the growing trends ofresearch and prototyping in the grid-interactive efficient buildings domain,characteristics beyond simple algorithm forecast accuracy are important indetermining the algorithm s true utility for the smart buildings. These othercharacteristics include the overall design of the deployed architecture and theoperational efficiency of the forecasting system. In this work, we present adeep-learning-based load forecasting system that predicts the building load atone-hour interval for 18 hours in the future. We also present the challengesassociated with the real-time deployment of such systems as well as theresearch opportunities presented by a fully functional forecasting system thathas been developed within the National Renewable Energy Laboratory sIntelligent Campus program.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×