MATEC Web Conf.
Volume 211, 2018The 14th International Conference on Vibration Engineering and Technology of Machinery (VETOMAC XIV)
|Number of page(s)||6|
|Section||ND: Nonlinear Dynamics, Chaos and Control of Elastic Structures; TP6: Condition monitoring, tip-timing, experimental techniques|
|Published online||10 October 2018|
Effect of noise on support vector machine based fault diagnosis of IM using vibration and current signatures
Department of Mechanical Engineering, SGSITS Indore, MP,
b Department of Mechanical Engineering, Indian Institute of Technology Guwahati, Guwahati, Assam, 781039, India
This paper analyzes the effect of noise on support vector machine (SVM) based fault diagnosis of IM (IM). For this, a number of mechanical (bearing fault, unbalanced rotor, bowed rotor and misaligned rotor) and electrical faults (broken rotor bar, stator winding fault with two severity levels and phase unbalance with two severity levels) of IM are considered here. The vibration and current signals are used here for the diagnosis. Different experiments were performed in order to generate these signals at various operating condition of IM (Speed and Load). Time domain feature are then extracted from the raw vibration and current signals obtained from the experiments. Then, the noise are added in the raw signals and the same features are extracted from this corrupted signals. The features from the original and corrupted signals are used to feed the classifier. The one-versus-one multiclass SVM are used here to perform multi-fault diagnosis. The comparative analysis of performance of the SVM classifier using data with and without noise is presented.
© The Authors, published by EDP Sciences, 2018
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.