Fault diagnosis system of rotating machines using continuous wavelet transform and Artificial Neural Network

In this paper using a machine with a motor configuration that is connected with 3 discs. Performance of a machine can be known by analyzing the vibrations that occur in the machine. Vibration that occurs on the machine may be normal or abnormal. Abnormal vibrations on a machine can cause severe damage. This abnormal vibration can be caused by the mass distribution of rotation no longer exists in the centerline. This technique of identifying vibrations can use a combination of Continuous Wavelet Transform (CWT) and Artificial Neural Network (ANN) methods. The vibration signal is sampled to be transformed using CWT, so the data of Continuous Wavelet Coefficient (CWC) is obtained. The Feature Extraction method is used to extract the Continuous Wavelet Transform data into several types. Root Mean Square (RMS), Kurtosis, and Power Spectrum Density (PSD) are Feature Extraction types used as Artificial Neural Network inputs to identify abnormal vibrations in the machine. The Artificial Neural Network (ANN) intelligently classifies the fault from machine vibrations. CWT and ANN combinations are able to classify the damage by 99.72% accuracy.