RBF neural network based PI pitch controller for a class of 5-MW wind turbines using particle swarm optimization...

Area of Science:

• Renewable Energy Engineering
• Control Systems Theory
• Artificial Intelligence in Engineering

Background:

• Wind turbine pitch control is crucial for stable operation and power regulation.
• Proportional-Integral (PI) controllers are widely used for their simplicity and industrial applicability.
• Optimizing PI controller gains is challenging due to system complexities and uncertainties.

Purpose of the Study:

• To propose a novel PI controller for collective pitch control (CPC) of a 5-MW wind turbine.
• To leverage artificial intelligence, specifically radial basis function (RBF) neural networks and particle swarm optimization (PSO), for optimal controller gain determination.
• To demonstrate a method that bypasses the need for detailed system nonlinearities and uncertainties.

Main Methods:

• Development of a radial basis function (RBF) neural network-based PI controller.
• Utilizing particle swarm optimization (PSO) to generate an optimal dataset for training the RBF neural network.
• Simulation-based performance evaluation of the proposed controller on a 5-MW wind turbine model.

Main Results:

• The proposed RBF neural network-PI controller demonstrated satisfactory performance in controlling the pitch angle.
• The PSO-optimized training dataset enabled effective learning for the RBF neural network.
• The controller achieved effective collective pitch control without explicit modeling of system complexities.

Conclusions:

• The RBF neural network combined with PSO offers an effective approach for optimizing PI controllers in wind turbines.
• This method provides a robust solution for collective pitch control, handling system uncertainties.
• The proposed controller shows significant potential for practical implementation in wind energy systems.