Droop coefficient placements for grid-side energy storage considering nodal frequency constraints under large disturbances

The high -voltage cross -regional power injections threaten the power systems under high renewable penetrations. The system operators should keep the frequency nadirs within restrictions after possible large disturbances to guarantee frequency stability. However, the heterogeneous characteristics, including the network topology, differences in disturbance locations, regulation units, and operating modes, make the nodal frequency response behaviors different after large disturbances. Therefore, the center of inertia frequency is unsuitable for evaluating the frequency stability of the entire power system. At the same time, the primary regulations from energy storage with proper droop settings are expected to solve the power grid's frequency stability problems. This paper focuses on the droop coefficient placements for grid -side energy storage, considering nodal frequency constraints. We use data -driven methods, i.e., alternative support vector machine trees (ASVMTREE), to extract the rules of different droop placement strategies' influences on nodal frequency stability. Then, We optimize the droop coefficient of grid -side energy storage for typical operating modes. Finally, we verify the method on modified IEEE 39 and 118 -bus test systems to show its effectiveness.