Fault Diagnosis of Helical Gear Box using Large Margin K-Nearest Neighbors Classifier using Sound Signals

TL;DR

This paper presents a machine learning method using Large Margin K-Nearest Neighbors to diagnose faults in helical gearboxes based on sound signals, overcoming limitations of traditional FFT analysis for non-stationary signals.

Contribution

It introduces a novel fault diagnosis approach utilizing sound signals and a Large Margin KNN classifier, with feature selection via J48 decision tree, for improved accuracy.

Findings

Effective fault classification achieved with sound signals.

Feature selection improves classifier performance.

Parameter effects on accuracy are analyzed.

Abstract

Gear drives are one of the most widely used transmission system in many machinery. Sound signals of a rotating machine contain the dynamic information about its health conditions. Not much information available in the literature reporting suitability of sound signals for fault diagnosis applications. Maximum numbers of literature are based on FFT (Fast Fourier Transform) analysis and have its own limitations with non-stationary signals like the ones from gears. In this paper, attempt has been made in using sound signals acquired from gears in good and simulated faulty conditions for the purpose of fault diagnosis through a machine learning approach. The descriptive statistical features were extracted from the acquired sound signals and the predominant features were selected using J48 decision tree technique. The selected features were then used for classification using Large Margin K-nearest neighbor approach. The paper also discusses the effect of various parameters on classification accuracy.