Haar and Symlet Discrete Wavelete Transform for Identification Misalignment on Three Phase Induction Motor Using Energy Level and Feature Extraction

Abstract

Currently induction motors are widely used in industry due to strong construction, high efficiency, and cheap maintenance. Machine maintenance is needed to prolong the life of the induction motor. As studied, bearing faults may account for 42% -50% of all motor failures. In general it is due to manufacturing faults, lack of lubrication, and installation errors. Misalignment of motor is one of the installation errors. This paper is concerned to simulation of discrete wavelet transform for identifying misalignment in induction motor. Modelling of motor operation is introduced in this paper as normal operation and three variations of misalignment. For this task, haar and symlet discrete wavelet transform in first level until fifth level is used to extract vibration signal of motor into high frequency of signal. Then, energy signal and other signal extraction gotten from high frequency signal is evaluated to analysis condition of motor. The results show that symlet discrete wavelet transform at the first to fourth level and combined with feature extraction of sum type can identify normal motor and misalignment motor conditions well.