Spatio-Temporal Forecasting of renewable energies

Overview

This project aims to develop short-term (up to 24 hours ahead) forecasting models that take into account the spatial as well as temporal information in wind farms and solar farms. Such models are useful for operational planning in farms and stabilising the network.

Research activities

• Gathering and analysing data for wind and solar power forecasting from Australian farms.
• Embedding Information such as the geographical location, spatial correlations and distances for developing forecasting models.
• Developing a novel method that will outperform the state of the art in terms of accuracy and biasness over the very short term or short-term horizon (up to one day ahead).
• Developing novel models for hierarchical forecasting by considering geospatial features for forecasts and customised loss functions that incorporate this information into the learning process.

Outcomes

• A machine learning model for forecasting wind/solar power.
• Publications in reputable journals and conferences.

Skills and experience

Familiarity with machine learning, familiarity with statistical analysis.

Scholarships

You may be eligible to apply for a research scholarship.

Explore our research scholarships

Keywords

• Forecasting
• Machine learning
• Renewable energies

Contact

Contact the supervisor for more information.