This is a project which is currently making use of HPC facilities at Newcastle University. It is active.
For further information about this project, please contact:
This project will use high-performance computing to develop and evaluate deep-learning–based control strategies for low-carbon building operations under uncertainty. Large-scale simulations of building energy systems will be combined with historical and forecast data on electricity carbon intensity, prices, weather, and demand to train and test deep reinforcement learning models. The research involves parallel training of neural networks, hyperparameter sweeps, scenario-based uncertainty analysis, and multi-objective optimization across carbon emissions, energy costs, and peak demand. HPC resources are used to accelerate model training, enable extensive experimentation across buildings and operating conditions, and support high-resolution, data-driven decision-making for demand-side energy management.
The project will use GPU-accelerated deep learning frameworks such as PyTorch and TensorFlow to develop and train deep reinforcement learning models for building energy management. Large-scale data processing and numerical computations will be performed using Python scientific libraries including NumPy, SciPy, and pandas. HPC GPUs will be essential for efficient neural network training, parallel hyperparameter optimization, and large-scale scenario evaluation under uncertainty. The workflow will leverage batch job scheduling and parallel processing on the HPC system to enable reproducible, high-throughput experimentation and multi-objective optimization across carbon emissions, energy cost, and peak demand.