An analysis of a model base signal for trouble detections [in electric machines]

In this paper, the authors consider a signal model based on frequency modulation. They present an algorithm for the detection of optimal parameters for approximation of signals. As a result, by evaluating errors, signals sampled by a sufficiently large sampling frequency can be approximated by the proposed algorithm. The authors use signals observed from the bearings of rotating machines as input signals, and show that detected parameters are useful for trouble detection.