Enhanced DC-Link Stability in DFIG-Based Wind Energy System Using PI and ANN Controllers

Authors: Pala Ganesh, Thadi Eswari, Paila Chaturya, Gemmeli Siva, Kundari Bharat kumar ,K. Sathish Kumar

Abstract 

Doubly Fed Induction Generator (DFIG)-based wind energy conversion systems are widely used due to their variable-speed operation and reduced converter rating. However, maintaining stable DC-link voltage under varying wind speeds, nonlinear loads, and grid disturbances is a critical challenge. This paper presents a comprehensive control strategy for a grid-connected DFIG wind system with battery energy storage (BES) at the DC-link. An improved phase-locked loop (PLL) is implemented to suppress DC-offset in grid voltages. Both conventional proportional-integral (PI) and artificial neural network (ANN)-based controllers are designed and comparatively evaluated for DC-link voltage regulation. The ANN controller demonstrates superior dynamic response, reduced overshoot, faster settling time, and better robustness against parameter variations and nonlinear operating conditions. The rotor-side converter (RSC) ensures unity power factor and maximum power extraction using tip-speed ratio-based MPPT, while the grid-side converter (GSC) regulates constant grid power export/import based on averaged wind power and battery state-of-charge (SOC). The system is modeled and simulated in MATLAB/Simulink, and performance is validated on a laboratory prototype. Results confirm improved DC-link stability, enhanced power quality, and reliable operation under wind turbulence, step changes, and unbalanced nonlinear loads.

Keywords: DFIG, Wind Energy Conversion System, DC-Link Voltage Control, Artificial Neural Network, Phase-Locked Loop, Battery Energy Storage, Maximum Power Point Tracking