Research Article

Wind Turbine Anomaly Identification Based on Improved Deep Belief Network with SCADA Data

Table 2

Prediction model performance.

Model classificationInput parameterIndex
RMSEMAEMAPE (%)

WT generator prediction modelGNT. bearing T. (B) (°C)1.621.554.57
WT gearbox prediction modelGearbox oil sump T. (°C)0.750.581.05
Whole WT prediction modelWind power (kW)21.4413.965.4