CompactNet: a light-weight deep learning framework for smart intrusive load monitoring

Minh H. Phan, Queen Nguyen*, Son L. Phung, Wei Emma Zhang, Trung D. Vo, Quan Z. Sheng

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Energy load monitoring via smart plugs or smart sockets has become more and more popular. Various studies have been undertaken to monitor energy consumption of household appliances and analyze the collected power data to obtain useful insights on consumers’ behaviors. The main challenge in load monitoring is to automatically recognize appliances in real time since the existing energy disaggregation process is time-consuming and labour-intensive. Although several deep learning models can achieve high accuracy on appliance classification, they usually consume large memory, hence not suitable for resources-constrained IoT devices. To resolve the issue, we demonstrate in this paper, for the first time, a novel framework named Smart Intrusive Load Monitoring based on a compact network (CompactNet), which is able to determine appliance types in real time. Specifically, our method distills the knowledge of an ensemble of large deep networks to a much more compact network. Our CompactNet accurately classifies various types of appliances, but its size is reduced by approximately eight times, making it possible to be deployed on edge IoT sensors for appliance recognition.

Original languageEnglish
Pages (from-to)25181-25189
Number of pages9
JournalIEEE Sensors Journal
Volume21
Issue number22
Early online date9 Jun 2021
DOIs
Publication statusPublished - 15 Nov 2021

Keywords

  • artificial intelligence
  • Feature extraction
  • Hidden Markov models
  • Home appliances
  • Internet of Things
  • intrusive load monitoring
  • knowledge distillation
  • Knowledge engineering
  • Load modeling
  • Monitoring
  • Scalable deep learning
  • smart sockets
  • Sockets

Fingerprint

Dive into the research topics of 'CompactNet: a light-weight deep learning framework for smart intrusive load monitoring'. Together they form a unique fingerprint.

Cite this