Advanced power control of a variable speed wind turbine based on a doubly fed induction generator using field-oriented control with fuzzy and neural controllers

Due to the rising prevalence of wind power in electrical power systems, wind farms are today required to participate actively in grid operation by an appropriate generation control. The objective of this article is to improve the energy quality of variable speed wind turbines by utilizing a doubly fed induction generator (DFIG) and developing a comprehensive control method. In pursuit of this objective, our focus has been on the DFIG rotor's ability to regulate active and reactive power using field-oriented control to a set point ordered by the wind farm control system. We harnessed the potential of three distinct types of controllers: the proportional-integral (PI), fuzzy logic controller (FLC), and a neural network-based controller (NNC), each designed for precise vector control of the DFIG. The PI controller parameters are adjusted based on the machine's datasheet, utilizing guideline-based tuning rules found in conventional synthesis methods. However, it is important to note that this tuning process demands significant computational time and a thorough understanding of the DFIM parameter set. To address this limitation, the conventional PI controller for power control is substituted with a FLC. The FLC lacks a rigorous mathematical foundation for stability analysis and optimization. This makes it difficult to guarantee certain system performance characteristics and stability. To enhance the control strategy's performance even further, we substituted the fuzzy logic controller with a neural network-based controller (NNC). The simulation results unequivocally demonstrate the superior performance of the NNC over both the FLC and the conventional PI controllers, particularly in terms of dynamic responsiveness and the system's resilience in accurately tracking power reference values. To evaluate the system's performance under different control strategies, we conducted simulations using the MATLAB/Simulink software.