9 Long-Term Automatic Tracking of the Modal Parameters of an Offshore Wind Turbine Drivetrain System in Standstill Condition 93 0.01 -0.01 0.005 -0.005 0 0 Acceleration , m/s2 100 200 300 Time, s 400 500 600 a c b Fig. 9.1 (a) The structure of the instrumented wind turbine; (b) Locations of the accelerometers on the drivetrain unit; (c) A typical measured acceleration signals 9.3 The Automatic Tracking Approach of the Drivetrain’s Modal Behavior In [14], a short-term automatic tracking of some of the drivetrain modes in a frequency band of 2–15 Hz was introduced, and it was shown that there are 10 physical vibration modes can be found in that frequency band. Therefore, in this paper and by means of the automatic mode tracking approach introduced in [14], those 10 physical modes of the drivetrain system will be tracked but over a longer period. The 3 steps of the used automatic mode tracking are illustrated in Fig. 9.2. In the first step, the time-domain data is loaded and transformed to the frequency-domain by calculating the power spectrum using the Periodogram approach [15, 16]. In the Periodogram approach, the time-domain data is divided into some overlapped blocks, and for each block the discrete Fourier transform (DFT) is calculated to obtain the spectrum. To reduce the leakage effects when applying the DFT, a Hanning window can be applied to each time-data block before the calculation of the DFT. The final power spectrum is taken as the average spectrum over the different processed time blocks as follows: b SYoYo .!k/ D 1 Nb Nb X bD1 Yw ob .!k/Yw ob .!k/ 2R (9.1) b SYoYref .!k/ D 1 Nb Nb X bD1 Yw ob .!k/Yw Refb .!k/ 2C (9.2)
RkJQdWJsaXNoZXIy MTMzNzEzMQ==