Electrical engineering

Physics

Geology

Control engineering

Meteorology

Mathematics

Feed forward

Agronomy

Biology

Support vector machine

Statistics

Machine learning

Engineering

Magnet

Estimator

Artificial intelligence

Wind speed

Controller (irrigation)

Control (management)

Permanent magnet synchronous generator

Pitch angle

Variable speed wind turbine

Turbine

Pitch control

Computer science

Feedforward neural network

Wind power

Control theory (sociology)

Artificial neural network

Mechanical engineering

Geophysics

Particle swarm optimization

IEEE Transactions on Energy Conversion

Although wind power has gained tremendous development in recent years, how to achieve mechanical loads optimization to extend service life-time of wind turbines is still a hot and challenging topic. In this study, artificial intelligence and advanced control techniques are combined to approach this objective for variable-speed wind turbine systems operating in high-speed region. First, the real-time information of effective wind speed is extracted and predicted via support vector regression (SVR) by exploiting data stream acquired online. Optimization of the support vector regression's parameters is completed by the particle swarm optimization algorithm. Subsequently, the predicted wind speed is routed to a novel feedforward mechanism designed to build a nonlinear relationship between wind speed and pitch angle. Additionally, an uncertainty and disturbance estimator (UDE) based feedback controller is implemented to deal with the model uncertainties and external disturbances. Both loads optimization and rotor speed/generator power regulation are achieved via strict math analysis. Finally, extensive comparison studies between the proposed scheme and traditional pitch angle controllers are conducted on GH bladed platform to verify the feasibility and effectiveness of the proposed scheme.

Hybrid Intelligent Feedforward-Feedback Pitch Control for VSWT With Predicted Wind Speed