Călin Ciufudean, Constantin Filote, Adrian Graur, Daniel Popescu, Camelia Petrescu
Diagnosability of Electric Machines Using Randomized Modulation Supply
This paper proposes a diagnosis method for electric machine faults by measuring the harmonic spectrum of stator current. In our assumption an electrical machine is excited with a randomized modulation of switching in power converters for reducing filtering requirements and acoustic noise in motor drive applications. Therefore, we propose a new approach for modelling the randomized modulation is the availability of an explicit control of time-domain performance, in addition to the possibility of shaping the power spectrum of signal of interest. The tool used here for investigating and classifying the harmonic decomposition of disturbances is the Least-Squares Prony method. Further approaches of this new method, as well as an example are also discussed.