Skip to Content
Find More Like This
Return to Search

A Proactive Learning Framework for Non-Intrusive Load Monitoring

National Renewable Energy Laboratory

Contact NREL About This Technology


Technology Marketing Summary

Non-intrusive load monitoring (NILM) is an emerging class of load-disaggregation technology that promises to replace expensive submetering. NILM systems perform analysis on whole-building data taken at the main power panel to accurately estimate the energy consumption of individual appliances. Existing NILM technology requires extensive user inputs which result in inconvenience and low rate of adoption.

 

Description

Engineers at the National Renewable Energy Laboratory (NREL) have developed software to overcome these barriers by enabling effective human-machine interaction. There is little incremental cost for implementation since the software can leverage the existing NILM infrastructure and perform data analytics in the cloud or on existing mobile platforms. The software augments existing NILM technology by involving the user in the loop. The software is built upon the framework of proactive learning, an advanced machine learning method that interactively queries the user for the class label information without assuming the user is infallible or indefatigable. Unlike most existing NILM systems which heuristically request user inputs, the software only needs the minimally sufficient information from the user to build a compact and yet highly representative load signature library. Thus, the software substantially reduces the burden on the user, improves the performance of a NILM system with limited user inputs, and overcomes the key barriers to wide adoption of NILM technology. The software has been shown to reduce the user input by up to 90% while still achieving similar performance compared to existing NILM systems.

 

Benefits
  • Proactive learning approach significantly reduces the burden on the user
  • Improved performance of NILM technology with limited user inputs
  • Flexibility of the proactive learning framework ensures the seamless incorporation of the solution into existing NILM systems

 

Applications and Industries
  • Non-Intrusive Load Monitoring (NILM)
  • Electrical Load Disaggregation
  • Home Automation
  • Energy Audit
  • Fault Detection and Diagnosis
  • Advanced Metering Infrastructure

 

Technology Status
Technology IDDevelopment StageAvailabilityPublishedLast Updated
NREL SWR 16-19DevelopmentAvailable03/25/201603/25/2016

Contact NREL About This Technology

To: Doreen Molk<doreen.molk@nrel.gov>