1
Return

Lightweight WaveNet-Enhanced Deformable Attention Model for Appliance-Level Building Energy Monitoring and Disaggregation

delete2026-08-05
delete0
PRE
AI
M
Mohammed Ayub
E
El-Sayed M El-Alfy *
DOI：10.1016/j.enbuild.2026.118055delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• Proposes a lightweight NILM model for accurate appliance-level energy monitoring in buildings. • Applies dilated convolution and deformable attention to capture key load dynamics efficiently. • Introduces an appliance-weighted loss to improve detection of low-consumption building devices. • Validates the approach using two public datasets under multi- and single-appliance conditions. • Demonstrates improved disaggregation accuracy that enables smarter building energy management.
Keywords:
Smart Building
Energy Monitoring
Energy Disaggregation
Deformable Attention
Time Series
Deep Learning
Huber Loss
Encoder

Journal

Energy and Buildings cover
Energy and Buildings
IF:
7.1
Papers:
1.5W
Citations:
6.8W

Organization

I
Information and Computer Science Department
Scholars:
28
Papers: 12
Citations: 0
Cited Papers

Cited Papers

Citing Papers

Citing Papers