Abstract :

Reliable Condition Monitoring Of Rotating Machinery Is Essential To Ensure Uninterrupted Industrial Operations And To Avoid Catastrophic Failures. Gearboxes, Being Critical Transmission Elements, Are Highly Prone To Faults Such As Tooth Breakage, Wear, And Misalignment, Which Significantly Alter Their Vibration Patterns. This Paper Introduces A Data-driven Methodology For Gearbox Fault Detection Using Vibration Signal Analysis. The Approach Is Based On Normalized Least Mean Square (NLMS) Adaptive Filtering, Which Is Applied To Raw Vibration Signals To Extract Predictive Error Statistics. The Standard Deviation Of NLMS Error, Along With Time-domain Features Such As Signal Power, Root Mean Square (RMS), Kurtosis, And Crest Factor, Are Computed To Characterize The Underlying Condition Of The Gearbox. These Features Provide Discriminative Insight Into Differences Between Healthy And Faulty Signals: While Healthy Gears Often Exhibit Higher Power But More Predictable Dynamics, Faulty Gears Are Characterized By Unpredictable Patterns And Distinctive Impulsive Features. A Support Vector Machine (SVM) Classifier Is Trained Using These Features To Distinguish Between Healthy And Broken-tooth Gear Conditions. Experimental Evaluation On Vibration Datasets With Approximately 88,000 Samples Per File Demonstrates That The Proposed Framework Achieves High Classification Accuracy, With Fault Signatures Showing Consistent Deviations In Power And Error-based Metrics. Furthermore, The Approach Is Integrated Into A MATLAB-based Graphical User Interface (GUI) That Allows Users To Upload Signals, Perform Feature Extraction, Visualize Periodicity Through Autocorrelation, And Obtain Real-time Fault Classification Results. The Contributions Of This Work Include: (1) A Novel Integration Of NLMS Error-based Features With Statistical Descriptors For Gearbox Fault Detection, (2) An Effective SVM Classification Pipeline For Binary Fault Diagnosis, And (3) A Practical MATLAB GUI Implementation For Realtime Monitoring. This Methodology Has Strong Potential For Industrial Deployment And Can Be Extended To Multifault Scenarios In Future Work

Published:

24-4-2026

Issue:

Vol. 26 No. 4 (2026)

Page Nos:

2991 - 2999

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite