Derivative-Free Data-Enabled Optimal Tracking (DF-DeeOT) Controller for a PV Grid-Connected Inverter

The growing demand for the adoption of renewable energy has introduced significant challenges to the stability and reliability of the power grid. Rapid advances in data-driven techniques, coupled with the increasing capabilities of embedded computing systems, present new opportunities to overcome various modeling and control limitations in modern electricity grids. In this paper, we propose a Data-Enabled Optimal Tracking (DeeOT) algorithm for a PV Grid-Connected Inverter System (GCIS). In addition, we propose a Derivative-Free version of DeeOT (DF-DeeOT) that does not require gradient estimation and does not assume the availability of initial stable control policies.We assess the performance of the proposed techniques via extensive MATLAB/Simulink simulations and laboratory-scale physical GCIS. The results suggest that the proposed algorithms are efficient in achieving optimal tracking under various grid conditions and outperform the existing controller.