Simple Model Predictive Control of High Power Direct-Driven PMSG Wind Energy Systems

Abstract

In this paper, a simple model predictive controller with constant switching frequency is designed to determine switching states of machine and grid NPCs. The proposed controller is designed based on modulated model predictive control principle to reduce computational burden of the algorithm. Time delay and weight factors in cost functions of the conventional model predictive control are eliminated to simplify the controller design. Also, space vector modulation algorithm is introduced to fix switching frequency of the converter. Machine side NPC is controlled to regulate the electromagnetic torque, stator flux and capacitors voltage of NPC while grid side NPC is controlled to regulate the active power, reactive power injected to the grid and DC-link voltage. The performance of the proposed controller is evaluated using 3MW/3kV generation system model. Simulation results of the proposed algorithm are good in spite of using low switching frequency.