Dynamic optimization-based method for determining the flexibility potential at vertical system interconnections

The integration of distributed energy resources (DERs) is reshaping distribution grids and changing their role in the power system which requires more efficient coordination between grid operators. The available power flexibility in a distribution grid can be aggregated as active and reactive power flexibility region at the interconnection point between transmission system and the distribution system. The Feasible Operation Region (FOR) represents the available flexibility on the P – Q plane at this interconnection point. Existing methods for approximating the FOR include different random sampling approaches, which generate valid power flow points but become computationally inefficient for large distribution grids. Other approaches are optimization-based which solve a series of optimization problems to approximate the FOR boundary, however, such approaches face challenges as DERs increase due to non-convexity of the problem. This paper proposes a novel heuristic-based dynamic sampling strategy using the REvol evolutionary algorithm to walk along the FOR boundary efficiently. This method begins by finding an initial boundary point and dynamically adapts its search direction using domain knowledge and population reuse to improve overall performance. Preliminary results on benchmark grids show promise, and the proposed dynamic method is expected to outperform random sampling as the complexity of the grid increases.