Matrix pencil method and fft based embedded condition monitoring of rotating machine

Rotating machines are an integral part of large electrical power machinery in most of the industries. Any degradation or outages in the rotating electric machinery can result in significant losses in productivity. It is critical to monitor the equipment for any degradation's so that it can serve as an early warning for adequate maintenance activities and repair. Prior research and field studies have indicated that the rotating machines have a particular type of signal structure during the initial start-up transient. This signal structure can be modelled by an exponentially damped sinusoidal signal during initial start-up. The major focus of this project is to take advantage of this a prior knowledge of signal structure to study the effect of degradation in two main signal parameters: frequency components of transient signal and the associated exponential damping factor.

In this work, a data acquisition system is first used to sample the transient signal. This signal is analysed on an embedded TMS320 based DSP core platform. Both FFT and MPM estimators have been implemented to study various signals captured by the data-acquisition system. For measuring the frequency of operation, both FFT and MPM were employed whereas the damping coefficient is calculated using MPM method. Initial implementation and validation was conducted in MATLAB and Simulink. Once validated, the code was ported on to TMS320 platform. The results show good agreement with expected results for both simulated and real-time data. The real-time data from AC water pumps which have rotating motors built-in were collected and analysed.

It is found that the frequency estimation performance of the MPM matches that of the FFT for synthetic and real time data. The evaluation of the condition of machines based on the damping coefficient using MPM matched with the actual condition of the machine.