Severity Invariant Feature Selection for Machine Health Monitoring

Vibration signals used for abnormality detection in machine health monitoring (MHM) suffer from significant variation in the patterns with fault severity. This variation results in overlap among the features extracted against different fault types and causes severe degradation in fault detection accuracy. This paper identifies a newfangled problem originated by severity variant features and mitigates this impact by using appropriate feature selection based on Fisher linear discriminant (FLD) and Bhattacharyya distance (BCD) to enhance fault classification accuracy. In order to validate the performance of the proposed scheme, training and testing data are obtained from different severity levels. To capture the non-stationary behavior of vibration signal, robust tools such as wavelet transform (WT) for time-frequency analysis is employed. Simulation studies show that the proposed scheme ensures good fault diagnostic accuracy even if training and testing data belong to different severity levels. Copyright (C) 2011 Praise Worthy Prize S.r.l. - All rights reserved.