Fixed Switching Frequency Model Predictive Controller for Doubly-Grounded Five-Level Photovoltaic Inverter

Abstract

Recently, doubly-grounded (DG) multilevel inverters (MIs) have found wide utilization in grid-connected photovoltaic (PV) applications. The direct connection between the grid neutral with the negative terminal of the PV side results in the total elimination of leakage current components. Existing classical controllers suffer from cascaded control structures, which make the design process more complex. Model predictive control (MPC) methods are advantageous over classical controllers at handling multiple objectives simultaneously. However, the variable switching frequency and weighting factor design issues represent the main challenges for the MPC design in PV inverters. Therefore, this paper presents a fixed switching frequency MPC method for DG-MIs in PV applications based on using optimized switching sequences. The proposed method can achieve fixed switching frequency operation with reduced dv/dt values. Moreover, the proposed method eliminates the weighting factors from the cost function, which makes the design process more simple. Simulation results are performed for the proposed method on five-level DG-MI for PV applications.