Formulation of Switching Instant for Improved Dynamic Performance in the Predictive Current Control Technique

Abstract

This paper analyzes the dynamic performance of the predictive current control (PCC) scheme for a single-phase grid-connected (SPGC) voltage source inverter (VSI). VSI performance and its operating conditions, such as steady-state and dynamic-state, have been analyzed. Conventionally a linear current error (eL)-based control technique is used, which has poorer current tracking during dynamic conditions. In this paper, a novel control technique has been proposed that utilizes a non-linear current error (eNL) based on the instantaneous grid voltage and DC-link voltage to calculate the switching instant, which gives better tracking performance during the dynamic condition. A mathematical model of digitally controlled PCC predicts the load current for all switching states of the inverter. Calculations of these predicted current values are done by using a cost function. The switching state with minimum cost function is selected and applied to the inverter for improved dynamic performance. The effect of parameter variations, such as load current, DC-link voltage etc., has been investigated. The performance of the proposed method has been tested at different sampling frequencies. The effectiveness of the PCC technique is simulated in specialized power electronics software and experimentally validated by a proto-type model using an FPGA controller.