Selecting time-frequency representations for detecting rotor faults in BLDC motors operating under rapidly varying operating conditions

Electric motors work continuously under operating conditions that rapidly vary with time. Motor diagnostics in a non-stationary environment is challenging due to the need for sophisticated signal processing techniques. In this paper, the use of quadratic TFRs is presented as a solution for the diagnostics of rotor faults in brushless DC (BLDC) motors operating under constantly changing load and speed conditions. Four time-frequency representations are considered in this paper-short-time Fourier transform (STFT), Wigner-Ville distribution (WVD), Choi-Williams distribution (CWD), and the Zhao-Atlas marks distribution (ZAM). The drawbacks of these distributions and methods to overcome them are also presented. The TFRs are implemented in real-time and their load computations are compared in order to study their suitability for implementation in a commercial system.