Data Fusion Based on an Iterative Learning Algorithm for Fault Detection in Wind Turbine Pitch Control Systems

Area of Science:

• Engineering
• Control Systems
• Renewable Energy

Background:

• Wind turbine pitch actuator failures can lead to significant operational issues and downtime.
• Accurate and timely detection of these failures is crucial for maintaining energy production and reducing maintenance costs.

Purpose of the Study:

• To propose a novel iterative learning algorithm for sensor data fusion.
• To detect failures in wind turbine hydraulic pitch actuators.

Main Methods:

• The proposed algorithm is developed based on iterative learning control and Lyapunov's theories.
• Numerical experiments were conducted using a standard wind turbine hydraulic pitch actuator model.

Main Results:

• The algorithm demonstrated effectiveness in detecting common pitch actuator faults.
• Simulations included faults such as high oil content, hydraulic leaks, and pump wear.

Conclusions:

• The developed iterative learning algorithm offers a promising approach for early detection of wind turbine pitch actuator failures.
• This method enhances the reliability and efficiency of wind turbine operations through advanced diagnostics.