Glasgow-based technology company Smarter Grid Solutions (SGS) has employed two forms of artificial
intelligence to improve the forecasts needed to run flexibility services. As part of its contribution to Western
Power Distribution’s Electricity Flexibility and Forecasting Systems (EFFS) project, SGS used “long shortterm
memory” and “extreme gradient boosting” forecasting techniques to make predictions that are more effective than classic statistical methods. The work for the EFFS project has already gained interest from
other distribution network operators, including Scottish Power Energy Networks and Scottish and Southern
Electricity Networks, which are looking to implement similar techniques within their own projects.
