Efficient model predictive power control with online inductance estimation for photovoltaic inverters

In this paper, a computationally efficient finite-set model predictive power control for grid-connected photovoltaic systems combined with a novel online finite-set model inductance estimation technique is proposed. The proposed control scheme overcomes the well-known challenges associated with predictive control in power electronics applications, which are: high model dependency and short sampling periods. The reference voltage vector (VV) of the grid-connected inverter that will enhance the desired power flow can be computed analytically with the knowledge of the reference and actual measured power values. Based on its location in the α–β reference frame, a finite set of three candidates instead of seven is evaluated to select the optimal VV. Furthermore, the performance of the proposed scheme is compared with the traditional finite-set model predictive power control, voltage oriented control with PI controllers, lookup table direct power control. Finally, the novel online inductance estimation technique is described and compared with unscented Kalman filter.