Event Detection and Feature Extraction for Home Appliances Based on Matrix Profile 


Vol. 44,  No. 6, pp. 1140-1146, Jun.  2019
10.7840/kics.2019.44.6.1140


PDF
  Abstract

As per capita power consumption per person continues to increase, the interest in NILM (Non-Intrusive Load Monitoring) is growing as a way of saving energy. NILM is a technology to reduce energy consumption by inducing user"s energy saving through real-time energy consumption monitoring. In this study, we propose a method to increase the event recognition rate by applying the Matrix Profile algorithm. This algorithm detects transient state interval and extracts the signature of a single home appliances. To increase the detection rate of transient event, current data from the three household appliances were collected using a high frequency data acquisition system (DAQ). To evaluate the performance of the proposed method, we collected noise data with a power size similar to that of household appliances. In addition, we apply a resampling method to the current and the Matrix Profile data in the transient state to implement a home appliance classification model based on CNN and GRU. Experiments have shown that it is possible to classify between a noise data having a power level similar to the specific home appliance and a home appliance, and it can be classified only by current data without loss of real-time information.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Cite this article

[IEEE Style]

H. Kang and H. Kim, "Event Detection and Feature Extraction for Home Appliances Based on Matrix Profile," The Journal of Korean Institute of Communications and Information Sciences, vol. 44, no. 6, pp. 1140-1146, 2019. DOI: 10.7840/kics.2019.44.6.1140.

[ACM Style]

Hyo-eun Kang and Ho-won Kim. 2019. Event Detection and Feature Extraction for Home Appliances Based on Matrix Profile. The Journal of Korean Institute of Communications and Information Sciences, 44, 6, (2019), 1140-1146. DOI: 10.7840/kics.2019.44.6.1140.

[KICS Style]

Hyo-eun Kang and Ho-won Kim, "Event Detection and Feature Extraction for Home Appliances Based on Matrix Profile," The Journal of Korean Institute of Communications and Information Sciences, vol. 44, no. 6, pp. 1140-1146, 6. 2019. (https://doi.org/10.7840/kics.2019.44.6.1140)

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880