Clustering Data-Driven Local Control Schemes in Active Distribution Grids

Controllable distributed energy resources (DERs) in active distribution grids (ADGs) provide operational flexibility to system operators, thereby, offering the means address various challenges. Existing local controllers for these are communication-free, robust, and cheap, but with suboptimal performance compared centralized approaches that heavily rely on monitoring communication. Data-driven controls can bridge gap by providing customized designed from historical data, offline optimization, machine learning methods. These emulate optimal behavior under expected operating conditions, without use of However, they exhibit high implementation overhead need individual programming DER controllers, especially when there many DERs or new units installed at a later stage. In this article, we propose clustering method decrease reducing into smaller set while still achieving performance. We show three-phase, unbalanced, low-voltage, network.