Short-term solar energy forecasting from sky video sequences

People

Supervisor

Research areas

Description

Global warming as well as climate change has become very critical issues worldwide. It has been one highly possible factor of causing the bushfires happened in US, South America and Australia. Using clean energy (such as solar or wind energy) is gradually adopted across the world to reduce the Co2 output to the atmosphere. In Australia, household solar panel becomes popular. However, it causes issues in integrating the solar energy into the existing power system (see https://www.abc.net.au/news/2019-12-01/rise-of-rooftop-solar-power-jeopardising-wa-energy-grid/11731452 ). Therefore, in this project, the goal is to use sky videos to forecast the solar power generation for up to 30 minutes. It would involve spatial-temporal encoding of video sequences and its applications in short-term solar energy forecasting. Students are expected to understand the 3D geometry, state-of-the-art spatial-temporal modeling in machine learning for solar energy forecasting.

 

Contact person: Miaomiao Liu (Miaomiao.liu@anu.edu.au)

Requirements

Programming in Python, Deep Learning, Pytorch, Project particularly for Master of Machine Learning and Computer Vision programme in ANU.

Background Literature

Computer Science or Engineering Students

Keywords

Computer Vision, Machine Learning

Updated:  1 June 2019/Responsible Officer:  Head of School/Page Contact:  CECS Marketing