Smart Energy Grids
Dr Simon Tindemans works in the area of smart grids and electricity control. The energy domain has the same potential for “flash crashes” – or wide-ranging failures – that have been seen in the financial domain. Simon’s work focuses on ways to avert or respond to these, as well as ways to use machine learning to continually adapt smart electricity provision. In particular, the work performed as part of the Low Carbon London initiative examines the mechanisms by which smart pricing and electricity provision can modify human behaviour to encourage sustainability.
Adaptative pricing models
In the future, this research could lead to new models about how people will respond and react into issues around energy provision and smart pricing, and how to can ensure reliable provision given potential changes in the behavioural model.
Machine learning to test energy grid reliability
High-intensity energy systems have high-reliability requirements (and, even if these aren’t met, they’re at least close to being met). This can make testing and validation difficult because a lot of resources are spent on tests that don’t show anything – i.e. in looking at circumstances in which the electricity grid doesn’t fail. Simon is interested in using machine learning to zero in on the scenarios whereas electricity grids do fail so that it is possible to test and validate activities in those areas.