Physics Informed Neural Networks for Control Oriented Thermal Modeling of Buildings
@article{Gokhale2022PhysicsIN,
title={Physics Informed Neural Networks for Control Oriented Thermal Modeling of Buildings},
author={Gargya Gokhale and Bert Claessens and Chris Develder},
journal={ArXiv},
year={2022},
volume={abs/2111.12066}
}Figures and Tables from this paper
One Citation
Physically Consistent Neural Networks for building thermal modeling: theory and analysis
- Computer ScienceArXiv
- 2021
This work presents a novel physics-informed NN architecture, dubbed Physically Consistent NN (PCNN), which only requires past operational data and no engineering overhead, including prior knowledge in a linear module running in parallel to a classical NN.
References
SHOWING 1-10 OF 52 REFERENCES
Physics-constrained Deep Learning of Multi-zone Building Thermal Dynamics
- EngineeringEnergy and Buildings
- 2021
Optimal temperature control of intermittently heated buildings using Model Predictive Control: Part I – Building modeling
- Engineering
- 2012
Physics-informed linear regression is competitive with two Machine Learning methods in residential building MPC
- EngineeringApplied Energy
- 2022
Deep reinforcement learning to optimise indoor temperature control and heating energy consumption in buildings
- Engineering
- 2020
Model predictive control strategy applied to different types of building for space heating
- EngineeringApplied Energy
- 2018
Experimental Demonstration of Frequency Regulation by Commercial Buildings—Part I: Modeling and Hierarchical Control Design
- EngineeringIEEE Transactions on Smart Grid
- 2018
It is demonstrated that commercial buildings can track a frequency regulation signal with high accuracy and minimal occupant discomfort in a realistic environment, and it is shown that buildings can determine the reserve capacity and baseline power a priori, and identify the optimal tradeoff between frequency regulation and energy efficiency.