Robust PLL Algorithm for Three-Phase Grid-Connected Converters

The estimation of the grid angle is a key factor in the control of many power electronics applications using grid-connected three-phase converters. The knowledge of the grid angle is required for different goals such as the synchronization of the voltage or current waveforms produced by the converter to the utility grid, the power flow management and the reduction of the transients generated on grid failures or on the reconnection of the converter. Generally, the grid voltages are polluted with harmonics and affected by amplitude unbalances; thus, oscillation on the grid angle signal can lead to an improper synchronization between the converter outputs and the grid voltages. To improve the synchronization, the paper proposes a robust PLL structure employing a prediction-correction filter based on a second-order linear and time-invariant observation model. Such filter is able to provide an accurate tracking of the grid voltage angle also in critical operating conditions. The paper describes also the implementation on a low cost fixed-point DSP controller and a comparison, obtained by simulation, with traditional filters. Finally, some significant experimental results validate the performances of the proposed approach.