Dataset for Paper "An Embedded Deep Learning NILM System A Year-long Field Study in Real Houses"

Citation Author(s):
Simone
Mari
University of L'Aquila
Giovanni
Bucci
University of L'Aquila
Fabrizio
Ciancetta
University of L'Aquila
Edoardo
Fiorucci
University of L'Aquila
Andrea
Fioravanti
University of L'Aquila
Submitted by:
Simone Mari
Last updated:
Sun, 08/27/2023 - 18:01
DOI:
10.21227/rhmz-7c39
Data Format:
*.csv
License:
Creative Commons Attribution
325 Views
Categories:
Power and Energy
Keywords:
Convolutional neural network (CNN), deep learning (DL), Edge Deployment, Energy Disaggregation, Nonintrusive Load Monitoring (NILM).
0
0 ratings - Please login to submit your rating.
ACCESS DATASET CITE SHARE/EMBED

Abstract 

This dataset contains electrical measurements collected in the context of the paper "An Embedded Deep Learning NILM System: A Year-long Field Study in Real Houses".

Nonintrusive load monitoring (NILM) systems are used to identify the energy consumption patterns of individual devices in an electrical system, but broadening their market availability is a significant challenge. In this paper, a NILM system using edge processing is proposed, in which energy consumption data are processed directly on the device installed at the monitored facility. Specifically, it uses a sequence-to-point approach based on a convolutional neural network  implemented on an Arm Cortex-M7 microcontroller. This paper also reports the results of an extensive testing phase. The NILM system was installed in two real houses in central Italy to evaluate its installation and potential application in real-world scenarios. This study presents a promising solution that enables the widespread adoption of NILM systems by reducing their implementation cost and complexity and addresses the privacy concerns associated with cloud-based data processing. The results of our real-world testing provide compelling evidence of the potential of the proposed NILM system in various applications, including smart homes, building automation, and industrial energy management.

Instructions: 

README

Description

This dataset contains electrical measurements recorded in an apartment located in central Italy (referred to as House 1 in the paper “An Embedded Deep Learning NILM System: A Year-long Field Study in Real Houses”) . The measurements were collected over a period of 6 months with a sampling interval of 8 seconds. This sampling interval was chosen in order to evaluate the transfer learning capability of NILM systems trained on the REFIT dataset. The data represent aggregate and appliance-level quantities, enabling the evaluation of the performance of NILM systems. The measurements are collected and processed using a certified class 0.2 single-phase onboard meter, which ensures the reliability of the metrological data.

Information on Electrical Quatities

The recorded electrical quantities are divided into four channels:

  • Channel 1: Aggregate electrical quantities of the entire apartment
  • Channel 2: Electrical quantities involved in the operation of the fridge
  • Channel 3: Electrical quantities involved in the operation of the washing machine
  • Channel 4: Electrical quantities involved in the operation of the dishwasher

Data Format

The dataset is stored in a CSV (Comma-Separated Values) file, with the following columns:

  • Time: Date and time of the recording
  • Vrms_X: RMS voltage of channel X
  • Arms_X: RMS current of channel X
  • Watt_X: Active power of channel X
  • VAr_X: Reactive power of channel X

Where X can be 1, 2, 3 or 4, corresponding to the above channels.

Dataset Files

LOGIN TO ACCESS DATASET FILES

QUESTIONS?