Condition monitoring of slow-rotating components in wind turbine drivetrains using novel signal processing approaches based on the instantaneous angular speed signal (FWOTM995)

Recently, it was shown that the direct analysis of the instantaneous angular speed (IAS) signal allows for the detection of fault frequencies of mechanical components in rotating machinery. This finding proved that it is possible to detect mechanical faults through the torsional deflection of a rotating shaft. In contrast, conventional vibration monitoring typically assumes transverse deflection of the machine or component housing. Direct measurements of the IAS of a shaft are however not subjected to the same transfer path hindrances as vibration monitoring and have the potential to detect damage much earlier on.

These signals typically contain much less complex mixtures of interfering signals since it is measured closer to the potential fault source. To the author's knowledge, no research has been done yet that goes beyond a Fourier spectrum analysis of the IAS forming a significant scientific knowledge gap.
This project explores more advanced methodologies to process the IAS signal for
fault identification and diagnostics. The development of a novel concept for a multi-delay frequency domain filter is proposed for the goal of signal separation. Additionally, an investigation of the IAS fault signal deformations will enable the development of novel signal filtering approaches for increasing the signal-to-noise ratio of potential faults. Lastly, the developed algorithms will be optimized and validated using experimental encoder data from a wind turbine
gearbox.

Keywords:Condition Monitoring, Signal Processing, Instantaneous angular speed

Disciplines:Signal processing not elsewhere classified, Analogue and digital signal processing, Acoustics, noise and vibration engineering