OR

Browse By

Machine Learning, Data Analytics, and Knowledge Management for Microgrid Data (2017–2020)

Abstract:
Australia¿s electricity supply is centralised and fossil fuel based, leading to electricity overpricing and environmental non-sustainability. This has attracted wide attention from both the Queensland Government and the industry to invest in the renewable energy sector. With tremendous data being collected by sensors equipped with substations, transformers, PV panels, batteries, smart meters, and various smart home appliances, it is expected that much of the higher-level development for creating the grid of the future will rely on effective big data analytics and machine learning heuristics. This project will focus on developing large-scale machine learning methods to manage supply and demand in a smart grid with data provided by the Queensland Government and the industry partner. Combining realtime sensory data, historical power generation data, power consumption data, battery data, and weather data, this project aims at constructing a comprehensive knowledge graph to capture patterns and relationships for all entities in a microgrid ecosystem. Tools and applications will be developed based on the knowledge graph to automatically and dynamically predict short-term and long-term energy supply, demand, and storage. It will also develop a block-chain based system that can facilitate fully distributed, reliable and secure power trading in any islanded microgrid in Queensland.
Grant type:
Advance Queensland Research Fellowships
Funded by:
Queensland Government Department of Science, Information Technology and Innovation