A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Deep Learning Models for Future Occupancy Prediction in Residential Buildings
This paper contributes to the occupancy prediction problem by developing state-of-the-art deep learning models. The occupancy prediction problem is addressed from two different viewpoints: multi-label classification and a sequence-to-sequence time-series analysis using encoder-decoder architectures. The following deep learning algorithms are employed in this study to construct occupancy models: multi-layer perceptron (MLP), recurrent neural networks, long-short term memory (LSTM), gated recurrent units (GRU), and bidirectional LSTMs. The performance of these models is evaluated and compared in terms of accuracy and computational speed. The results demonstrate that addressing this problem using MLP models provides the best performance for short-term predictions, while for predictions more than 90 min ahead, GRU results in the highest accuracy. It is also demonstrated that the accuracy of the deep learning models can be approximated as a function of the occupancy index with an MAE of 0.014.
Deep Learning Models for Future Occupancy Prediction in Residential Buildings
This paper contributes to the occupancy prediction problem by developing state-of-the-art deep learning models. The occupancy prediction problem is addressed from two different viewpoints: multi-label classification and a sequence-to-sequence time-series analysis using encoder-decoder architectures. The following deep learning algorithms are employed in this study to construct occupancy models: multi-layer perceptron (MLP), recurrent neural networks, long-short term memory (LSTM), gated recurrent units (GRU), and bidirectional LSTMs. The performance of these models is evaluated and compared in terms of accuracy and computational speed. The results demonstrate that addressing this problem using MLP models provides the best performance for short-term predictions, while for predictions more than 90 min ahead, GRU results in the highest accuracy. It is also demonstrated that the accuracy of the deep learning models can be approximated as a function of the occupancy index with an MAE of 0.014.
Deep Learning Models for Future Occupancy Prediction in Residential Buildings
Environ Sci Eng
Wang, Liangzhu Leon (editor) / Ge, Hua (editor) / Zhai, Zhiqiang John (editor) / Qi, Dahai (editor) / Ouf, Mohamed (editor) / Sun, Chanjuan (editor) / Wang, Dengjia (editor) / Esrafilian-Najafabadi, Mohammad (author) / Babahaji, Mina (author) / Haghighat, Fariborz (author)
International Conference on Building Energy and Environment ; 2022
Proceedings of the 5th International Conference on Building Energy and Environment ; Chapter: 106 ; 999-1006
2023-09-05
8 pages
Article/Chapter (Book)
Electronic Resource
English
A new modeling approach for short-term prediction of occupancy in residential buildings
| British Library Online Contents | 2017