Primary supervisor
Terrence MakCo-supervisors
The research project aims to investigate:
- Multi-Model Fusion with Deep Neural Networks for Future Energy Systems (Smart Grid).
Future energy systems are envisioned to be running decentrally with full automatic control, high proportion of renewable energy (e.g., wind & solar), and abundant storage facilities. With many types of renewable energy sources are weather and climate dependent, accurate and timely prediction on reliability risks (e.g., loss of generation, voltage issues, and thermal limit violations) due to weather/climate are often necessary.
In power systems, numerical weather/climate data are often chosen to be used for crafting energy system prediction models due to its ease of storage, analysis, and simplicity in training. However, recent research in climate science also shows that a large category of datasets are images (e.g., radar / satellite image dataset), which can also be used to improve prediction for weather/climate-related forecasting. With majority of research in energy/power system focused on numerical weather/climate data, there are huge research potentials in exploring weather/climate data that are not natively numerical.
This project will explore various non-numerical weather/climate data to assist energy reliability prediction, and devise a multi-model fused neural network to effectively capture and predict reliability risks for future energy systems.
Student cohort
Required knowledge
Research activities
- Research and study machine learning techniques for nonlinear transportation systems
- Research and study power system applications
Coding activities
- Pytorch (Python based)
Skills/Knowledge required
- Good understanding on mathematical optimisation
- Prior experience in building deep learning models
- Prior experience in data cleaning and preprocessing