ABODE-Net: An Attention-based Deep Learning Model for Non-intrusive Building Occupancy Detection Using Smart Meter Data

TL;DR

ABODE-Net is a novel deep learning model with a parallel attention mechanism that effectively detects building occupancy from smart meter data, outperforming existing methods in accuracy.

Contribution

The paper introduces ABODE-Net, a deep learning model with a novel Parallel Attention block that improves occupancy detection accuracy using smart meter data.

Findings

ABODE-Net outperforms state-of-the-art models in occupancy detection accuracy.

The Parallel Attention block effectively captures important features for classification.

The model demonstrates robustness across different datasets.

Abstract

Occupancy information is useful for efficient energy management in the building sector. The massive high-resolution electrical power consumption data collected by smart meters in the advanced metering infrastructure (AMI) network make it possible to infer buildings' occupancy status in a non-intrusive way. In this paper, we propose a deep leaning model called ABODE-Net which employs a novel Parallel Attention (PA) block for building occupancy detection using smart meter data. The PA block combines the temporal, variable, and channel attention modules in a parallel way to signify important features for occupancy detection. We adopt two smart meter datasets widely used for building occupancy detection in our performance evaluation. A set of state-of-the-art shallow machine learning and deep learning models are included for performance comparison. The results show that ABODE-Net significantly outperforms other models in all experimental cases, which proves its validity as a solution for non-intrusive building occupancy detection.