Fault Detection of Wind Turbine Sensors Using Artificial Neural Networks

摘要: This paper proposes a method for sensor validation and fault detection in wind turbines. Ensuring validity of measurements is significant part overall condition monitoring as faults lead to incorrect results system’s state health. Although identifying abrupt failures sensors relatively straightforward, calibration drifts are more difficult detect. Therefore, isolation technique on the purpose measurement was developed. Temperature from Supervisory Control Data Acquisition system turbine were used this aim. Low output rate nonlinear characteristics drive necessity design an advanced algorithm. Artificial neural networks chosen considering their high performance environments. The demonstrate that proposed can effectively detect existence drift isolate exact with faulty behaviour.