CNN-LSTM architecture for predictive indoor temperature modeling

Indoor temperature modeling is a crucial part towards efficient Heating, Ventilation and Air Conditioning (HVAC) systems. Data-driven black-box approaches have been an attractive way to develop such models due to their unique feature of not requiring detailed knowledge about the target zone. However...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Building and environment Ročník 206; s. 108327
Hlavní autoři: Elmaz, Furkan, Eyckerman, Reinout, Casteels, Wim, Latré, Steven, Hellinckx, Peter
Médium: Journal Article
Jazyk:angličtina
Vydáno: Oxford Elsevier Ltd 01.12.2021
Elsevier BV
Témata:
Air conditioning
Air temperature
Artificial neural networks
Black-box modeling
Closed-loop prediction
CNN-LSTM
Computer architecture
Feature extraction
HVAC
Indoor temperature modeling
Long short-term memory
LSTM
MLP
Modelling
Multilayers
Neural networks
Predictions
Room temperature
Short term
MLP
Closed-loop prediction
HVAC
LSTM
CNN-LSTM
Indoor temperature modeling
Black-box modeling
ISSN:0360-1323, 1873-684X
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Indoor temperature modeling is a crucial part towards efficient Heating, Ventilation and Air Conditioning (HVAC) systems. Data-driven black-box approaches have been an attractive way to develop such models due to their unique feature of not requiring detailed knowledge about the target zone. However, the noisy and non-linear nature of the problem remains a bottleneck especially for long prediction horizons. In this paper, we introduce a Convolutional Neural Networks-Long Short Term Memory (CNN-LSTM) architecture to combine the exceptional feature extraction of convolutional layers with the Long Short Term Memory (LSTM)’s capability of learning sequential dependencies. We experimentally collected a dataset and compared three approaches: Multi-Layer Perceptron (MLP), LSTM and CNN-LSTM. Models are evaluated and compared with 1-30-60-120 min horizons with a closed-loop prediction scheme. The CNN-LSTM outperformed all other models for all prediction horizons and showed a better robustness against error accumulation. It managed to predict room temperature with R2>0.9 in a 120-min prediction horizon. •CNN-LSTM architecture is proposed for data-driven indoor temperature modeling.•Proposed architecture is compared to vanilla neural network and LSTM architectures.•Results are evaluated in 1-30-60-120 min prediction horizons.•Proposed architecture outperformed other techniques and showed superior stability.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0360-1323
1873-684X
DOI:10.1016/j.buildenv.2021.108327