英文摘要 |
Structural health monitoring (SHM) is of primary importance in wind turbine monitoring systems because the integral functionality of the turbine relies heavily on the robustness of the supporting structure. Traditional frequency domain tools limit the precise interpretation of the in situ vibration signals of an onshore wind turbine. In this study, we employ time-frequency analysis to identify the structural as well as mechanical dynamic characteristics of an onshore wind turbine (Zephyros Z72, 1.5 MW), based on the vibration signals of the supporting structure under parked (idling) and operational conditions. The dynamic signatures of the structure could be estimated from the vibration signals by system identification tools. However, several non-structurally dynamic patterns unexpectedly appeared in the time-frequency representation while analyzing the vibration signals by short-time Fourier transform (STFT). We thus employed STFT in combination with the concepts of mechanical vibration and rotor dynamics to resolve these implicit clues into explicit illustrations of probable events that occurred in the machinery, such as the run-up of the rotor, blade shadowing effects, and the subsequent beating phenomena over its supporting structure. |