Economic and environmental concerns coupled with government directives have led to significant interest in incorporating greater amounts of renewable energy (e.g., solar and wind energy) into the grid. The ongoing project to establish a smart microgrid at the Kuala Belalong Field Studies Centre (KBFSC) in Brunei is one such initiative. However, issues such as variability in demand and supply, large cost of storage, cuncertainty about the availability of wind and sunlight, and fast variations in renewable power pose barriers to its inclusion.
In order to address these challenges, we design novel algorithms for intelligent online generation scheduling and storage management. Specifically, our algorithms (a) allow us to compensate for the renewable energy when it is not available, (b) take into account the physical generator constraints, (c) adapt on the fly to uncertainty as short term predictions about the wind intensity and sunlight become available, and (d) manage local storage and distribution based on these short term predictions. Besides being theoretically near-optimal, our algorithms demonstrate excellent practical performance even in the worst-case scenarios.
More info:
Paper 1
Paper 2
Paper 3