Fault detection of wind turbine system using neural networks

Electricity produced from wind energy is one of the rapid growing power generation methods in the world. Kinetic energy in wind rotates the rotor blade in wind turbine system thus generating power. As the wind turbine system has many components, chance of fault development is more in the turbine system. System faults can lead to increased mechanical component degradation, severe reduction of asset performance, and a direct increase in annual maintenance costs. Possible faults in wind turbine system are blade angle asymmetry and yaw misalignment. This work focuses on the detection of faults in the wind turbine system using Artificial Neural Networks (NN). Modeling of turbine system is done by neural networks with the real time data. 600 samples are taken randomly and are used to train the neural network system; another 300 samples are taken for validation. Threshold limits are done by model error modeling method. For a fault free system, output should present inside the threshold limits. The best model is compared which precisely detect fault in the system.