Performance Analysis of Model Predictive Technique Based Combined Control for PMSG-Based Distributed Generation Unit

Abstract

In this article, we propose a predictive model technique based on combined control for distributed generation unit (DGU) driven by the uncontrolled prime mover. The DGU constitutes of a permanent magnet synchronous generator, along with back-to-back converter, and is operated to supply distributed load at fixed voltage and frequency. The model predictive control is developed by using two cost functions for the generator- and load-side converters simultaneously. The cost function for the generator-side converter is generated in the synchronously rotating reference frame, whereas the load side is developed in the stationary reference frame. The cost functions are developed to maintain the power flow from the generator at the desired level of demand throughout the operation. The proposed controller identifies the switching state of the converters without using the pulsewidth modulation modulator. The DGU is, thus, controlled to maintain the load voltage and frequency, irrespective of the uncontrolled variation of the load impedance and speed of the prime mover. Additionally, the controller is tested with the variation of the active power component of the load voltage and deviation in the dc-link voltage. The implementation of the proposed control is performed in the simulation environment successfully. The experimental results are presented to validate the theoretical performances in a scaled laboratory prototype.