Building energy consumption prediction based on Bayesian optimized LSTM model

Yiming Zhang; Jing Yang; Sen Zhou; Pengchun Li

doi:10.1117/12.3011544

7 December 2023 Building energy consumption prediction based on Bayesian optimized LSTM model

Yiming Zhang, Jing Yang, Sen Zhou, Pengchun Li

Proceedings Volume 12941, International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 2023); 129414T (2023) https://doi.org/10.1117/12.3011544
Event: Third International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 203), 2023, Yinchuan, China

Abstract

Due to the low accuracy of energy consumption prediction when the hyper-parameters of the long-term and short-term neural network model (LSTM) are experimentally determined. This paper proposes to optimize the hyper-parameters based on Bayesian algorithm and apply them to LSTM to construct a combined energy consumption prediction model. Seven strongly correlated factor data of a building are selected as input feature quantities to train the combined model. The prediction results are analyzed to verify the prediction accuracy of the model. And the ARIMA, RNN and LSTM models are compared and verified. The results show that the prediction accuracy of Bayesian optimized LSTM model reaches 97.2 %, so it has high reliability for building energy consumption prediction.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Yiming Zhang, Jing Yang, Sen Zhou, and Pengchun Li "Building energy consumption prediction based on Bayesian optimized LSTM model", Proc. SPIE 12941, International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 2023), 129414T (7 December 2023); https://doi.org/10.1117/12.3011544

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available