MODEL PREDICTIVE POWER CONTROL APPROACH FOR THREE-PHASE SINGLE-STAGE GRID-TIED PV MODULE-INTEGRATED CONVERTER

Abstract

This project is about the concept of the three-phase module-integrated converters (MICs) incorporated in grid-tied large-scale photovoltaic (PV) systems. The current-source converter (CSC) with dc voltage boost capability, namely single stage power conversion system, is proposed for three-phase PV MIC system. A model predictive scheme with low switching frequency is designed to control the proposed topology in such a way that provides a certain amount of active and reactive power in steady-state operation and also provides a proper ratio of reactive power under transient conditions to meet the low voltage ride through (LVRT) regulations. To predict the future behaviour of current control values and switching states, a discrete-time model of the MIC is developed in synchronous reference frame. It is demonstrated that the injected active and reactive power can be controlled using minimizing the cost function introduced in the predictive switching algorithm. The proposed structure is simulated in MATLAB/SIMULINK software. An experimental verification is provided to justify the performance of the proposed control method through a 300-VA laboratory prototype. The results verify the desired performance of the proposed control scheme for exchanging of both active and reactive powers between the PV MIC and the grid within different operating conditions.

Existing System

In the existing system, FCS-MPC method for the LVRT enhancement of the grid-tied VSC topology is used.

Proposed System

This proposed system develops the application of the single-stage CSC topology for the three-phase grid-tied PV MIC system. A predictive control strategy is applied to the CSC-based MIC to provide balanced grid currents under disturbances, simultaneously with suitable active and reactive power regulation, allowing to fully meet the LVRT requirement. For this purpose, the system modelling is first presented in continuous-time and then converted to discrete-time models for the purpose of digital implementation. The control objectives are expressed as a fitness function. During each sampling interval, the fitness function is minimized using the actual measurements and predicted values for given switching states, which are then applied to the CSC-based MIC directly. Finally, the simulation and experimental evaluation of the proposed controller for a 300-VA laboratory prototype is conducted.

Architecture

Block diagram of the proposed control system for CSC-based MIC.