EMD and ANN based intelligent model for bearing fault diagnosis

Issue title: Special Section: Intelligent tools and techniques for signals, machines and automation

Guest editors: Smriti Srivastava, Hasmat Malik and Rajneesh Sharma

Article type: Research Article

Authors: Shah, Arjun Kumar^a | Yadav, Ashish^a | Malik, H.^{b; *}

Affiliations: [a] Division of Manufacturing Process and Automation Engineering, NSIT, New Delhi, India | [b] Department of Electrical Engineering, IIT Delhi, New Delhi, India

Correspondence: [*] Corresponding author. H. Malik, Department of Electrical Engineering, IIT Delhi, New Delhi-110016, India. E-mail: [email protected].

Abstract: Rolling bearing is an important mechanical element therefore its condition monitoring is necessary to ensure the steadiness of industrial machineries. In this paper, the open source vibration data have been processed using advanced signal processing techniques such as EMD method to extract more symmetric waves (IMFs) out of non-linear and non-stationary vibration signals. In addition to this, statistical time-domain and frequency-domain features are calculated and then J48 Decision Tree Algorithm is used for feature selection. The processed input signals have been used for comparative study of five different types of Artificial Neural Network (ANN) classifiers. The performance characteristics of MLP, PNN, GRNN, RBF and LVQ are shown in results and discussion section.

Keywords: Rolling bearing, signal processing, IMFs, empirical mode decomposition (EMD), fault detection, artificial neural networks (ANNs), MLPs, PNN, GRNN, RBF, LVQ

DOI: 10.3233/JIFS-169821

Journal: Journal of Intelligent & Fuzzy Systems, vol. 35, no. 5, pp. 5391-5402, 2018