14th IEEE International Conference on Control and Automation (ICCA), Alaska, United States Of America, 12 - 15 June 2018, pp.957-962
In this paper, we study identification of induction motors by using a recently developed covariance-based subspace algorithm from sound measurements. The sound data are collected by an array of five-microphones placed hemispherically around motors in a reverberant and noisy room. It is demonstrated that mechanical faults, frequently encountered in induction motors, are isolated by the proposed identification algorithm. The estimated acoustic noise spectra can be utilized in the development of preventive maintenance programs for induction motors in service.