Decentralized Data-Driven Voltage Control for Clustered PV Inverters with Local Deviation Priority

Abstract

This paper presents a decentralised, data-driven voltage control strategy designed to coordinate multiple photovoltaic (PV) inverters operating as a cluster, with a focus on mitigating local voltage deviations. The proposed framework is fully data-driven, obviating the need for prior knowledge of system parameters. A power-sharing mechanism is integrated to facilitate effective coordination among PV inverters within the cluster, dynamically adapting to both local and global voltage conditions. Simulation results on an actual distribution network confirm the method's effectiveness in maintaining voltage regulation at the point of common coupling (PCC) while ensuring that all local voltages remain within permissible operational limits. The approach exhibits robust adaptability to system variations, addressing challenges posed by high PV penetration and dynamic network changes. Numerical simulations conducted in MATLAB/Simulink highlight the method's potential to enhance grid stability and support the integration of renewable energy into modern distribution networks.