Journal
BUILDING SIMULATION
Volume 14, Issue 1, Pages 165-178Publisher
TSINGHUA UNIV PRESS
DOI: 10.1007/s12273-019-0597-2
Keywords
thermal model; building simulation; experimental validation; multi-layer perceptron; linear vs; non-linear; black-box vs; gray-box
Categories
Funding
- Institute for Data Valorization (IVADO)
Ask authors/readers for more resources
Thermal models of buildings are essential for forecasting energy use and improving mechanical system control. Autoregressive neural network models outperform gray-box and black-box linear models in simulating indoor temperatures with limited information.
Thermal models of buildings are helpful to forecast their energy use and to enhance the control of their mechanical systems. However, these models are building-specific and require a tedious, error-prone and time-consuming development effort relying on skilled building energy modelers. Compared to white-box and gray-box models, data-driven (black-box) models require less development time and a minimal amount of information about the building characteristics. In this paper, autoregressive neural network models are compared to gray-box and black-box linear models to simulate indoor temperatures. These models are trained, validated and compared to actual experimental data obtained for an existing commercial building in Montreal (QC, Canada) equipped with roof top units for air conditioning. Results show that neural networks mimic more accurately the thermal behavior of the building when limited information is available, compared to gray-box and black-box linear models. The gray-box model does not perform adequately due to its under-parameterized nature, while the linear models cannot capture non-linear phenomena such as radiative heat transfer and occupancy. Therefore, the neural network models outperform the alternative models in the presented application, reaching a coefficient of determination R-2 up to 0.824 and a root mean square error down to 1.11 degrees C, including the error propagation over time for a 1-week period with a 5-minute time-step. When considering a 50-hour time horizon, the best neural networks reach a much lower root mean square error of around 0.6 degrees C, which is suitable for applications such as model predictive control.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available