Presentations from WESC 2019.
Published June 18, 2019 | Version v1
Presentation Open

Probabilistic fault diagnostics using ensemble time-varying decision tree learning

Authors/Creators

  • 1. ETH Zurich

Description

Probabilistic fault diagnostics using ensemble time-varying decision tree learning on wind turbines. Simulations using Simulink and an embedded FAST aeroelastic model of the 5MW reference wind turbine. Error cases: (1) No Error, (2) Yaw Error,  Corrected, (3) Yaw Error, yaw actuator stuck, (4) Pitch angle sensor, stuck at constant value of 5 deg.

Files

Files (2.0 MB)

Additional details

Funding

European Commission
WINDMIL - Smart Monitoring, Inspection and Life-Cycle Assessment of Wind Turbines 679843
Swiss National Science Foundation
Smart Monitoring System for Inspection and Life-Cycle Assessment of Critical Infrastructure 206021_177012