Development and evaluation of forecasting models for solar power and electricity use over space and time

Time period: 2017-01-01 to 2020-12-31

Project leader: Joakim Munkhammar

Funder: Swedish Energy Agency

Type of award: Project grant

Total fundning: 4 055 173 SEK

This project aims to develop and evaluate forecasting methods for solar power and electricity use. The project will evaluate existing methods and based on that develop state-of-the-art methods for forecasting. These methods will be based on probabilistic forecasting, where for example artificial neural networks and ”deep learning” will be applied to forecasting of solar power and electricity use. Data from current and planned Uppsala county facilities will serve as basis for model validation. Based on this evaluation, decisions regarding Uppsala county policy for implementing metering systems for solar irradiance, solar power generation and electricity use will be made. The methods developed are intended to facilitate a basis for future design of smart grids.