Predictive Control of Common-Ground Five-Level PV Inverter Without Weighting Factors and Reduced Computational Burden

Mitigation of the leakage current components represents a critical issue in transformerless photovoltaic (PV) inverters. Recently, PV inverters with inherent common-ground (CG) connection are becoming promising proposals for nearly total elimination of leakage current components. However, developing multilevel output while injecting grid currents with high quality is a challenging task for classical control schemes. This article presents a reduced computational burden weighting factorless finite-control-set model-predictive control (WFL-FCSMPC) for the CG five-level single-phase switched capacitor cells inverter-based PV systems. The proposed reduced computational burden WFL-FCSMPC avoids the cascaded control loops of classical controllers. Thence, the problems related to adjusting weighting factors of standard finite control-set model-predictive control (FCSMPC) are eliminated by dividing the control objectives into the desired cost functions. Moreover, high computational burdens in classical FCSMPC methods are considerably reduced using the proposed WFL-FCSMPC, especially at a high number of levels. Therefore, robust and simplified design and implementation are obtained by applying the proposed WFL-FCSMPC while long tuning procedures of weighting factors are avoided. Simulation and experimental results are provided for verifying the superior performance of the proposed WFL-FCSMPC over standard FCSMPC and linear controllers.