Predictive control for static power converters working in wide frequency ranges

Distributed and nonconventional power generation sources could lead to a wide varying grid frequency, particularly in weak systems or islanding mode operation of micro-grids. This work studies a predictive control scheme for power converters that are directly connected to the AC network featuring large frequency variations. The proposed scheme considers a fixed number of samples per period independent of the network frequency, leading to a variable sampling time under varying frequency conditions. This feature is advantageous when frequency has an important operating range because it ensures a constant resolution of the measured system variables. As discrete predictive control requires the sampling time to make accurate predictions, a synchronization system is proposed that ensures a fixed number of samples per cycle through continuous adjustment of the sampling time. Simulations show good performance of the proposed algorithms.