Research Article

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

Table 1

Condition parameters of the overall health prediction model.

Condition parameters (SCADA data)
WT generator prediction modelWT gearbox prediction modelWhole WT prediction model

GNT. bearing T. (B)Gearbox oil sump T.Wind power
Phase current 1Gearbox input shaft T.GNT. speed
Phase current 2Gearbox inlet oil T.Rotor speed
Phase current 1Gearbox output shaft T.GNT. stator winding T. (U)
GNT. stator winding T. (U)GNT. speedGNT. stator winding T. (V)
GNT. stator winding T. (V)Rotor speedGNT. stator winding T. (W)
GNT. stator winding T. (W)GNT. stator winding T. (U)Phase current 1
Wind powerGNT. stator winding T. (V)Phase current 2
Wind speedGNT. stator winding T. (W)Phase current 3
Spindle impeller side T.Wind powerInverter torque feedback
GNT. bearing T. (B)GNT. speed feedback
Inverter torque feedbackWind speed
GNT. speed feedback