Adaptive Neuro-Fuzzy Control for Minimizing Submodule Capacitance in Modular Multilevel Converters for Wind Energy Systems

Abstract

The direct-drive permanent magnet synchronous generator (PMSG) and modular multilevel converter ‎‎(MMC)-based offshore DC wind turbine has emerged as a strong contender for the large-capacity wind energy ‎conversion system (WECS). Few technical challenges exist for MMC when used in medium potential WECS ‎for PMSG. The important one is the enormous submodule (SM) voltage fluctuation brought on by the PMSG ‎phase current, which has a high amplitude and low frequency. But because this topology's capacitor voltage is ‎floating, a larger capacitor is needed, which raises the project's cost. This research suggests a minimum ‎voltage ripple control across the capacitor for an MMC-based wind energy conversion systems. For wind ‎energy conversion systems based on MMC, a different control method known as CCVR (Constant ‎Capacitance Voltage Ripple) is suggested. The voltage ripple on the CM capacitor can be greatly reduced by ‎using this method since it allows for the inclusion of the propagating current's second harmonic component. ‎Because of this, smaller SM capacitors can be used. This paper introduces a constant capacitance voltage ripple (CCVR) control strategy and an ANFIS-based ‎controller to address these challenges, aiming to reduce SM capacitor size while maintaining optimal system ‎performance.‎‎