Model predictive voltage control of a single-phase inverter with output LC filter for stand-alone renewable energy systems

This work offers a simple and efficient model predictive voltage control strategy with a two-step prediction for improved output voltage control of single-phase inverter used for stand-alone renewable energy systems. The single-phase inverter with output LC filter is used to provide a sinusoidal output voltage, regardless of the arbitrary consumer load profiles. The suggested control algorithm uses the discrete-time model of the inverter and LC filter for two-step prediction of the output voltage for each possible switching state. Then, the control algorithm selects the switching state which minimizes the error between the output voltage and its reference without using any inner controllers or modulators. An accurate extrapolation methodology with reduced computational burden is used to predict the dynamic changes in the output voltage reference. The switching frequency reduction is also attained by considering a simple constraint in the suggested algorithm. Simulation results are presented to verify the feasibility and excellent performance of the suggested controller under linear and nonlinear load conditions, compared to a conventional voltage control strategy based on PI regulators and PWM. Furthermore, experimental tests using a dSPACE system with DS1104 controller have been done to confirm the simulations results.