Modeling and analysis of single-phase two-stage inverter for grid-connected photovoltaic application

Abstract

In order to extract the maximum amount of power from the PV modules and convert it into an appropriate form for grid integration, a power electronics interface is needed. This thesis focuses on analysis and modeling of the single-phase two-stage inverter for grid-connected PV application with special attention given to the harmonic issues. A DC/DC battery-integrated boost converter can serve as the front end in a two-stage PV inverter configuration. A steady-state analysis for this new type of converter is carried out. An application of this converter in distributed Maximum Power Point Tracking (MPPT) architecture is proposed and compared with the conventional boost converter. By adding energy storage devices, it has the potential to compensate for the intermittent attribute of the PV generators. It also yields a more stabilized DC-link voltage in the distributed MPPT application. This thesis also analyzes the harmonic distortions generated from a PV inverter, which mainly relate to the DC/AC stage of the PV inverter. The field measurements are carried out at the CSIRO micro-grid. The high total harmonic distortion phenomenon is found when the PV inverters operate at light load condition. The conventional model of a PV inverter does not contain any harmonic information. A linear model for the PV inverter is modified to analyze the harmonic formation process. The causes of harmonics are summarized for the PV inverter operating at different power levels. The constant DC-link voltage assumption is not necessarily true for the two-stage PV inverter. The model of the grid-connected PV inverter with feedback control becomes time-varying with consideration for the DC-link voltage ripple. A series of odd harmonics is identified to be caused by this double-line frequency DC-link voltage ripple. The transfer function cannot be derived for this model. It is a challenge to calculate the exact harmonic amplitudes. The closed-form solution for this time-varying model is derived based on an idea which is similar to the harmonic balance method. A guideline is provided to select the minimum size of capacitor for the DC-link without sacrificing the power quality. Simulation and experimental results are provided to verify the aforementioned analysis results. In addition, some suggestions for future work are given.