Power Optimization of Grid-Connected DFIG Wind Energy System Using RNN-PI Hybrid Controller in MATLAB/Simulink

The variable and random nature of wind creates serious problems in the control and optimization of wind energy systems. Conventional control techniques like the PID controller tend to be inadequate when handling the highly nonlinear and complex nature of wind turbine dynamics, causing poor performance and inefficient energy production. In order to overcome these drawbacks, in this paper, a Recurrent Neural Network (RNN) control approach is proposed for a wind energy conversion system. The system consists of a Doubly Fed Induction Generator (DFIG) that is fed into a three-phase electrical grid through a rectifier, a three-phase Voltage Source Inverter (VSI), and an LC filter. The RNN controller compares the active and reactive reference power (Pref and Qref) with the actual power fed to the grid (Pact and Qact) to generate the control signals for the DFIG output. These control signals are utilized to generate PWM pulses to control the VSI, facilitating the conversion of AC power from DC power and grid integration. Also, a Proportional-Integral (PI) controller is used to keep the necessary DC bus voltage level, allowing it to operate stably. The proposed hybrid control architecture leverages the learning capacity of the RNN and the stability of the PI controller to ensure enhanced energy efficiency and dynamic performance of the wind energy system.