Investigating Holarchies to Perform Capacity Allocation in Smart Grids

This thesis presents a holarchy, combined with two developed algorithms. These algorithms aim to mitigate grid congestion and stay within the bounds of the flexibility sources. The first algorithm is called the Curtailment and Reallocation algorithm. It takes the preference profiles into account of the grid users, as well as preference bounds (bounds in which the grid user desires to stay in between). With these preference profiles, it calculates a profile in which these bounds are met as much as possible, while preventing grid congestion. It does not take the limits of the flexibility sources into consideration, which can result in violations in that area. The second algorithm deals with that by trading capacity with other holons in a fully distributed way, such that they are able to fully mitigate these violations regarding flexibility without a third party being involved. In the end, the developed algorithms have shown to prioritize customer satisfaction at the cost of extra complexity. Compared to a state of the art algorithm, the developed algorithms adhere more to the preference profile, with a preference met of 83%, compared to 20.8% for the state of the art algorithm.