An Improved MPC Control of Grid-Connected Inverter - Application to PV System

Abstract

Model predictive control (MPC) of power converters is an efficient technique that has given recently very suitable performance. MPC can give effective tracking and regulation with fast dynamic response and robustness. The technique is based on a simple model to select which optimal voltage vector has to be applied in the next time interval without the need for a modular. The main drawback is that the switching frequency is variable and depends on the inverter operational point in terms of active and reactive power which is a typical case in photovoltaic (PV) applications. High switching frequency increases active losses in the converter and may have the risk of exceeding the allowable threshold of semiconductors. However, the very low switching frequency can degrade the produced power quality. This work proposes an improved MPC algorithm by taking into consideration the switching frequency evolution. The proposed controller looks for a trade-off between power quality and switching frequency when performing control. The proposed technique is implemented in MATLAB/SimPowerSystems, the good scenario is achieved with THD of 3.75% with a maximum switching frequency of 1200 Hz optimizing almost 90% from the maximum switching frequency when using standard MPC.