Non-Intrusive Load Monitoring of Household Devices Using a Hybrid Deep Learning Model through Convex Hull-Based Data Selection

Inoussa Laouali,Antonio Ruano,Saad Dosse Bennani,Maria Da Graça Ruano,Hakim El Fadili

doi:10.3390/en15031215

Inoussa Laouali, Antonio Ruano + Show 3 more

Open Access

https://doi.org/10.3390/en15031215

Copy DOI

Abstract

The availability of smart meters and IoT technology has opened new opportunities, ranging from monitoring electrical energy to extracting various types of information related to household occupancy, and with the frequency of usage of different appliances. Non-intrusive load monitoring (NILM) allows users to disaggregate the usage of each device in the house using the total aggregated power signals collected from a smart meter that is typically installed in the household. It enables the monitoring of domestic appliance use without the need to install individual sensors for each device, thus minimizing electrical system complexities and associated costs. This paper proposes an NILM framework based on low frequency power data using a convex hull data selection approach and hybrid deep learning architecture. It employs a sliding window of aggregated active and reactive powers sampled at 1 Hz. A randomized approximation convex hull data selection approach performs the selection of the most informative vertices of the real convex hull. The hybrid deep learning architecture is composed of two models: a classification model based on a convolutional neural network trained with a regression model based on a bidirectional long-term memory neural network. The results obtained on the test dataset demonstrate the effectiveness of the proposed approach, achieving F1 values ranging from 0.95 to 0.99 for the four devices considered and estimation accuracy values between 0.88 and 0.98. These results compare favorably with the performance of existing approaches.

Highlights

Introduction published maps and institutional affilThe world’s ever-increasing energy consumption and ongoing dependency on fossil fuel-based energies have created significant environmental concerns, in terms of carbon dioxide (CO2 ) emissions [1]
This paper proposes an Non-intrusive load monitoring (NILM) framework based on low frequency power data
A randomized approximation convex hull data selection approach using sliding windows of active and reactive power. It is based on the selection of the most informative vertices of the real convex hull and has the advantage of reducing memory needs of the learning algorithms

Summary

Introduction

The world’s ever-increasing energy consumption and ongoing dependency on fossil fuel-based energies have created significant environmental concerns, in terms of carbon dioxide (CO2 ) emissions [1]. Focusing on the European country hosting the current case-study household, Portugal, greenhouse gas (GHG) emissions increased by 13%. From 2014 to 2018 as a result of increased economic activity and a high proportion of fossil fuels in its energy supply [2]. Low-carbon economies have emerged as the focus of worldwide attention in order to minimize energy consumption and greenhouse gas emissions [3]. Share of renewable energy consumption in the energy mix, a minimum of 32.5% energy savings, and a 40% reduction in greenhouse gas emissions compared to 1990 levels [4]. Portugal was among the first countries in the world to set carbon neutrality goals for the iations

Methods

Results

Conclusion