Predicting indoor temperature distribution with low data dependency using recurrent neural networks

Accurately predicting indoor temperature distribution can provide valuable reference data, helping residents independently adjust HVAC equipment around them to ensure comfort while reducing unnecessary energy consumption. This study proposes a prediction framework composed of two neural networks, en...

Full description

Saved in:
Bibliographic Details
Main Authors: Jiahe Wang, Shohei Miyata, Keiichiro Taniguchi, Yasunori Akashi
Format: Article
Language:English
Published: Taylor & Francis Group 2025-03-01
Series:Journal of Asian Architecture and Building Engineering
Subjects:
Online Access:http://dx.doi.org/10.1080/13467581.2025.2474818
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849737556024885248
author Jiahe Wang
Shohei Miyata
Keiichiro Taniguchi
Yasunori Akashi
author_facet Jiahe Wang
Shohei Miyata
Keiichiro Taniguchi
Yasunori Akashi
author_sort Jiahe Wang
collection DOAJ
description Accurately predicting indoor temperature distribution can provide valuable reference data, helping residents independently adjust HVAC equipment around them to ensure comfort while reducing unnecessary energy consumption. This study proposes a prediction framework composed of two neural networks, enabling accurate indoor temperature distribution prediction with minimal training data in both temporal and spatial dimensions. The Dual-Stage Attention-Based Recurrent Neural Network calculates the importance ranking of feature values to enhance individual feature information and reduce training data volume, while Long Short-Term Memory is used to predict time-series features. From February to September 2022, 22 temperature sensors were installed in a target office to collect minute-by-minute indoor temperature data, which served as training and testing datasets. The results showed that, for short-term prediction, using data from five out of the 22 sensors collected over two weeks in winter (heating season) and summer (cooling season), the framework accurately predicted temperatures at the remaining 17 sensor locations, with a root mean squared error between 0.3 and 0.7. This study is significant for continuous indoor temperature prediction under low data volume conditions.
format Article
id doaj-art-41871bf688db40ab8a0227aa992aeb91
institution DOAJ
issn 1347-2852
language English
publishDate 2025-03-01
publisher Taylor & Francis Group
record_format Article
series Journal of Asian Architecture and Building Engineering
spelling doaj-art-41871bf688db40ab8a0227aa992aeb912025-08-20T03:06:53ZengTaylor & Francis GroupJournal of Asian Architecture and Building Engineering1347-28522025-03-010011410.1080/13467581.2025.24748182474818Predicting indoor temperature distribution with low data dependency using recurrent neural networksJiahe Wang0Shohei Miyata1Keiichiro Taniguchi2Yasunori Akashi3The University of TokyoThe University of TokyoThe University of TokyoThe University of TokyoAccurately predicting indoor temperature distribution can provide valuable reference data, helping residents independently adjust HVAC equipment around them to ensure comfort while reducing unnecessary energy consumption. This study proposes a prediction framework composed of two neural networks, enabling accurate indoor temperature distribution prediction with minimal training data in both temporal and spatial dimensions. The Dual-Stage Attention-Based Recurrent Neural Network calculates the importance ranking of feature values to enhance individual feature information and reduce training data volume, while Long Short-Term Memory is used to predict time-series features. From February to September 2022, 22 temperature sensors were installed in a target office to collect minute-by-minute indoor temperature data, which served as training and testing datasets. The results showed that, for short-term prediction, using data from five out of the 22 sensors collected over two weeks in winter (heating season) and summer (cooling season), the framework accurately predicted temperatures at the remaining 17 sensor locations, with a root mean squared error between 0.3 and 0.7. This study is significant for continuous indoor temperature prediction under low data volume conditions.http://dx.doi.org/10.1080/13467581.2025.2474818indoor temperature predictiontemperature distributionlstmattention mechanismspatial measured data
spellingShingle Jiahe Wang
Shohei Miyata
Keiichiro Taniguchi
Yasunori Akashi
Predicting indoor temperature distribution with low data dependency using recurrent neural networks
Journal of Asian Architecture and Building Engineering
indoor temperature prediction
temperature distribution
lstm
attention mechanism
spatial measured data
title Predicting indoor temperature distribution with low data dependency using recurrent neural networks
title_full Predicting indoor temperature distribution with low data dependency using recurrent neural networks
title_fullStr Predicting indoor temperature distribution with low data dependency using recurrent neural networks
title_full_unstemmed Predicting indoor temperature distribution with low data dependency using recurrent neural networks
title_short Predicting indoor temperature distribution with low data dependency using recurrent neural networks
title_sort predicting indoor temperature distribution with low data dependency using recurrent neural networks
topic indoor temperature prediction
temperature distribution
lstm
attention mechanism
spatial measured data
url http://dx.doi.org/10.1080/13467581.2025.2474818
work_keys_str_mv AT jiahewang predictingindoortemperaturedistributionwithlowdatadependencyusingrecurrentneuralnetworks
AT shoheimiyata predictingindoortemperaturedistributionwithlowdatadependencyusingrecurrentneuralnetworks
AT keiichirotaniguchi predictingindoortemperaturedistributionwithlowdatadependencyusingrecurrentneuralnetworks
AT yasunoriakashi predictingindoortemperaturedistributionwithlowdatadependencyusingrecurrentneuralnetworks