Three-phase four-leg active power filter based on nonlinear optimal predictive control

In this paper a nonlinear predictive control using optimal control is applied to the current control of a three-phase four-wire shunt active power filter. The control law has closed analytical form and on-line dynamic optimization is avoided, the control law is easy to implement. Three phase currents are sensed and then transformed to synchronous rotating reference frame, positive and negative load currents distortions are extracted and regarded as reference currents of the active power filter´s phase a, b and c, while zero sequence current in neutral wire is compensated by the fourth leg separately. A proportional-integral regulator is applied to the outer dc voltage loop and the compensated error is added to the inner loop d-axis current reference. Switching signals are generated through a 2-dimensional space vector pulse width modulator (SVPWM) for phase a, b, c, and through a triangular modulator for the fourth leg. The simulation results show that the APF based on the proposed method has good static and dynamic performance.