Comparative study of real-time stator voltage control in stand-alone doubly fed induction generators using PI, fuzzy logic and neural network approaches

This research focuses on the experimental comparative analysis of three techniques to optimize the control performance of the AC voltage generated by a doubly fed induction generator (DFIG) in a wind power generation system (WPGS) serving local consumers. Unlike grid-connected operation, rotor side control (RSC) during stand-alone operation focuses on maintaining stable voltage under variable conditions, such as wind-speed fluctuations and load variations. Traditional proportional-integral (PI) control is compared with intelligent control techniques, including fuzzy logic and artificial neural networks (ANN), to evaluate their effectiveness in enhancing performance. A real-time experimental setup was developed using a 3 kW DFIG and a dSPACE 1104 control platform to assess the performance of the three controllers. The experimental results under varying operating conditions reveal that the intelligent controllers significantly outperform the PI controller, particularly in terms of response time, robustness, and overshoot. The findings highlight the potential of intelligent controllers to enhance performance and stability in stand-alone DFIG applications.