APPLICATION OF PATTERN RECOGNITION TECHNIQUES FOR FAULT DETECTION OF CLUTCH RETAINER OF TRACTOR

Authors

  • Mostafa Bahrami Department of Mechanical Engineering of Agricultural Machinery, Faculty of Agriculture, Razi University, Kermanshah, Iran
  • Hossein Javadikia
  • Ebrahim Ebrahimi

DOI:

https://doi.org/10.3329/jme.v47i1.35356

Abstract

This study develops a technique based on pattern recognition for fault diagnosis of clutch retainer mechanism of MF285 tractor using the neural network. In this technique, time features and frequency domain features consist of Fast Fourier Transform (FFT) phase angle and Power Spectral Density (PSD) proposes to improve diagnosis ability. Three different cases, such as: normal condition, bearing wears and shaft wears were applied for signal processing. The data divides in two parts; in part one 70% data are dataset1 and in part two 30% for dataset2.
At first, the artificial neural networks (ANN) are trained by 60% dataset1 and validated by 20% dataset1 and tested by 20% dataset1. Then, to more test of the proposed model, the network using the datasets2 are simulated. The results indicate effective ability in accurate diagnosis of various clutch retainer mechanism of MF285 tractor faults using pattern recognition networks.

Downloads

Download data is not yet available.
Abstract
1254
PDF
1171

Downloads

Published

2018-05-01

How to Cite

Bahrami, M., Javadikia, H., & Ebrahimi, E. (2018). APPLICATION OF PATTERN RECOGNITION TECHNIQUES FOR FAULT DETECTION OF CLUTCH RETAINER OF TRACTOR. Journal of Mechanical Engineering, 47(1), 31–36. https://doi.org/10.3329/jme.v47i1.35356

Issue

Section

Articles