Multiple-teeth defect localization in geared systems using filtered acoustic spectrogram

Contactless health monitoring of machines is highly desirable in industrial setups where the environment inherently imposes restrictions on contact-based data acquisition. This motivates the use of acoustic signal as an effective alternative for condition monitoring of equipments located in such inaccessible environments. However, condition monitoring and fault diagnosis by processing acoustic signals still remains a challenge for researchers. The aim of the proposed work is to establish a robust technique of acoustic signal processing for detection and localization of multiple teeth defect in geared systems. Towards this, the present work proposes the use of time marginal integration (TMI) of the continuous wavelet transform (CWT) coefficients of the decomposed signal derived from an undecimated wavelet transform (UWT) of the raw acoustic signal. UWT, owing to its well established translation invariant property, is implemented on the raw data to extract the de-noised signal for further processing with CWT. The time-axis of the TMI graph is finally correlated to the angular displacement of the driver gear in order to locate the defective teeth and measure their relative positions. An artificial neural network (ANN) model using signal statistical parameters as neurons is proposed as a pre-check to identify the presence of any defect in the gears. In addition, the efficiency of UWT as a de-noising tool is reestablished through the accuracy improvement in ANN based identification. A synthetic signal is simulated to conceptualize and evaluate the effectiveness of the proposed method. Synthetic signal analysis also offers vital clues about the suitability of the biorthogonal 3.1 wavelet over Daubechies and Symlet wavelets in the proposed algorithm. The experimental validation of the proposed method is presented using a customized gear drive test setup by introducing gears with seeded defects in one or more of their teeth. Measurement of the angles between two or more damaged teeth with a high level of accuracy is shown to be possible using the proposed algorithm. Experiments reveal that acoustic signal analysis can be used as a suitable contactless alternative for precise gear defect identification and gear health monitoring.