Model Predictive Control Technique of Multilevel Inverter for PV Applications

Date

2018-05-02

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Renewable energy sources, such as solar, wind, hydro, and biofuels, continue to gain popularity as alternatives to the conventional generation system. The main unit in the renewable energy system is the power conditioning system (PCS). It is highly desirable to obtain higher efficiency, lower component cost, and high reliability for the PCS to decrease the levelized cost of energy. This suggests a need for new inverter configurations and controls optimization, which can achieve the aforementioned needs. To achieve these goals, this dissertation presents a modified multilevel inverter topology for grid-tied photovoltaic (PV) system to achieve a lower cost and higher efficiency comparing with the existing system. In addition, this dissertation will also focus on model predictive control (MPC) which controls the modified multilevel topology to regulate the injected power to the grid. A major requirement for the PCS is harvesting the maximum power from the PV. By incorporating MPC, the performance of the maximum power point tracking (MPPT) algorithm to accurately extract the maximum power is improved for multilevel DC-DC converter. Finally, this control technique is developed for the quasi-z-source inverter (qZSI) to accurately control the DC link voltage, input current, and produce a high quality grid injected current waveform compared with the conventional techniques. This dissertation presents a modified symmetrical and asymmetrical multilevel DC-link inverter (MLDCLI) topology with less power switches and gate drivers. In addition, the MPC technique is used to drive the modified and grid connected MLDCLI. The performance of the proposed topology with finite control set model predictive control (FCS-MPC) is verified by simulation and experimentally. Moreover, this dissertation introduces predictive control to achieve maximum power point for grid-tied PV system to quicken the response by predicting the error before the switching signal is applied to the converter. Using the modified technique ensures the iii system operates at maximum power point which is more economical. Thus, the proposed MPPT technique can extract more energy compared to the conventional MPPT techniques from the same amount of installed solar panel. In further detail, this dissertation proposes the FCS-MPC technique for the qZSI in PV system. In order to further improve the performance of the system, FCS-MPC with one step horizon prediction has been implemented and compared with the classical PI controller. The presented work shows the proposed control techniques outperform the ones of the conventional linear controllers for the same application. Finally, a new method of the parallel processing is presented to reduce the time processing for the MPC.

Description

Keywords

Model Predictive Control (MPC), Quasi-z-source Inverter (qZSI), Field programmable gate array (FPGA), DC-DC converter, Maximum power point tracking (MPPT), Potovoltaic (PV)

Citation