Sequence-to-Sequence Model with Transformer-based Attention Mechanism and Temporal Pooling for Non-Intrusive Load Monitoring

Abouzar Estebsari

doi:10.1109/EEEIC/ICPSEurope57605.2023.10194819.

Sequence-to-Sequence Model with Transformer-based Attention Mechanism and Temporal Pooling for Non-Intrusive Load Monitoring

Abouzar Estebsari

Civil and Building Services Engineering

Research output: Contribution to conference › Paper

Abstract

This paper presents a novel Sequence-to-Sequence (Seq2Seq) model based on a transformer-based attention mechanism and temporal pooling for Non-Intrusive Load Monitoring (NILM) of smart buildings. The paper aims to improve the accuracy of NILM by using a deep learning-based method. The proposed method uses a Seq2Seq model with a transformer-based attention mechanism to capture the long-term dependencies of NILM data. Additionally, temporal pooling is used to improve the model's accuracy by capturing both the steady-state and transient behavior of appliances. The paper evaluates the proposed method on a publicly available dataset and compares the results with other state-of-the-art NILM techniques. The results demonstrate that the proposed method outperforms the existing methods in terms of both accuracy and computational efficiency.

Original language	English
DOIs	https://doi.org/10.1109/EEEIC/ICPSEurope57605.2023.10194819.
Publication status	Published - 3 Aug 2023
Event	IEEE 23rd International Conference on Environment and Electrical Engineering (EEEIC) - Madrid, Spain Duration: 6 Jun 2023 → 9 Jun 2023 https://www.aconf.org/conf_187075.2023_IEEE_International_Conference_on_Environment_and_Electrical_Engineering_and_2023_IEEE_Industrial_and_Commercial_Power_Systems_Europe.html

Conference

Conference	IEEE 23rd International Conference on Environment and Electrical Engineering (EEEIC)
Country/Territory	Spain
City	Madrid
Period	6/06/23 → 9/06/23
Other	The aim of the conferences is to give the opportunity of a genuine and constructive dialogue among participants on the hot-topics and far-reaching challenges that engineers and scientists are called to face in the present days. The conference is so a precious chance to discuss recent developments and practical applications in crucial areas, such as sustainable and renewable energy production, energy storage, smart grids, microgrids, energy communities, smart buildings, energy conversion, sustainable transport systems, EMC control in lightning and grounding systems, novel materials and nanotechnology.
Internet address	https://www.aconf.org/conf_187075.2023_IEEE_International_Conference_on_Environment_and_Electrical_Engineering_and_2023_IEEE_Industrial_and_Commercial_Power_Systems_Europe.html

Access to Document

10.1109/EEEIC/ICPSEurope57605.2023.10194819.Licence: CC BY 4.0

Sequence-to-Sequence

Cite this

@conference{fb4aa831e795439c8eabf1b61b96aa7d,

title = "Sequence-to-Sequence Model with Transformer-based Attention Mechanism and Temporal Pooling for Non-Intrusive Load Monitoring",

abstract = "This paper presents a novel Sequence-to-Sequence (Seq2Seq) model based on a transformer-based attention mechanism and temporal pooling for Non-Intrusive Load Monitoring (NILM) of smart buildings. The paper aims to improve the accuracy of NILM by using a deep learning-based method. The proposed method uses a Seq2Seq model with a transformer-based attention mechanism to capture the long-term dependencies of NILM data. Additionally, temporal pooling is used to improve the model's accuracy by capturing both the steady-state and transient behavior of appliances. The paper evaluates the proposed method on a publicly available dataset and compares the results with other state-of-the-art NILM techniques. The results demonstrate that the proposed method outperforms the existing methods in terms of both accuracy and computational efficiency.",

author = "Abouzar Estebsari",

year = "2023",

month = aug,

day = "3",

doi = "10.1109/EEEIC/ICPSEurope57605.2023.10194819.",

language = "English",

note = "IEEE 23rd International Conference on Environment and Electrical Engineering (EEEIC) ; Conference date: 06-06-2023 Through 09-06-2023",

url = "https://www.aconf.org/conf_187075.2023_IEEE_International_Conference_on_Environment_and_Electrical_Engineering_and_2023_IEEE_Industrial_and_Commercial_Power_Systems_Europe.html",

}

TY - CONF

T1 - Sequence-to-Sequence Model with Transformer-based Attention Mechanism and Temporal Pooling for Non-Intrusive Load Monitoring

AU - Estebsari, Abouzar

PY - 2023/8/3

Y1 - 2023/8/3

N2 - This paper presents a novel Sequence-to-Sequence (Seq2Seq) model based on a transformer-based attention mechanism and temporal pooling for Non-Intrusive Load Monitoring (NILM) of smart buildings. The paper aims to improve the accuracy of NILM by using a deep learning-based method. The proposed method uses a Seq2Seq model with a transformer-based attention mechanism to capture the long-term dependencies of NILM data. Additionally, temporal pooling is used to improve the model's accuracy by capturing both the steady-state and transient behavior of appliances. The paper evaluates the proposed method on a publicly available dataset and compares the results with other state-of-the-art NILM techniques. The results demonstrate that the proposed method outperforms the existing methods in terms of both accuracy and computational efficiency.

AB - This paper presents a novel Sequence-to-Sequence (Seq2Seq) model based on a transformer-based attention mechanism and temporal pooling for Non-Intrusive Load Monitoring (NILM) of smart buildings. The paper aims to improve the accuracy of NILM by using a deep learning-based method. The proposed method uses a Seq2Seq model with a transformer-based attention mechanism to capture the long-term dependencies of NILM data. Additionally, temporal pooling is used to improve the model's accuracy by capturing both the steady-state and transient behavior of appliances. The paper evaluates the proposed method on a publicly available dataset and compares the results with other state-of-the-art NILM techniques. The results demonstrate that the proposed method outperforms the existing methods in terms of both accuracy and computational efficiency.

U2 - 10.1109/EEEIC/ICPSEurope57605.2023.10194819.

DO - 10.1109/EEEIC/ICPSEurope57605.2023.10194819.

M3 - Paper

T2 - IEEE 23rd International Conference on Environment and Electrical Engineering (EEEIC)

Y2 - 6 June 2023 through 9 June 2023

ER -

Sequence-to-Sequence Model with Transformer-based Attention Mechanism and Temporal Pooling for Non-Intrusive Load Monitoring

Abstract

Conference

Access to Document

Fingerprint

Cite this