A comparative study of Naïve Bayes classifier and Bayes Net classifier for fault diagnosis of automobile hydraulic brake system. (2015)

• Record Type: Journal Article
• Authors: Jegadeeshwaran, R.; Sugumaran, V.
• Is Part Of: International journal of decision support systems. Volume 1:Number 3(2015)
• Page Start: 247
• Page End: 267
• Publication Date: 2015
• Subjects: decision tree -- statistical features -- feature extraction -- naive Bayes classifier -- Bayes net classifier -- feature classification -- fault diagnosis -- brake faults -- automobile industry -- hydraulic brakes -- automotive braking systems -- condition monitoring -- continuous monitoring -- machine learning -- vibration analysis -- vehicle vibration; Decision making -- Data processing -- Periodicals; 658.40302856312
• Languages: English
• ISSNs: 2050-6988

Abstract

In an automobile, brake system must be reliable and effective. Any failure in the braking system that impacts the ability to retard a vehicle's motion will have an immediate or frequent catastrophic effect on the safety of the vehicle. Thus, the role of brake systems is critical and condition monitoring is required. Particularly, vibration-based continuous monitoring and analysis using machine learning approaches are gaining momentum. This study is one such attempt to perform fault diagnosis of hydraulic brake system by vibration analysis. In this paper, the performance of Bayes net and Naïve Bayes algorithms for fault diagnosis is presented. A hydraulic brake system test rig was fabricated. The vibration signals were acquired using a piezoelectric transducer. The statistical parameters are extracted and the good features that discriminate different faulty condition were identified. With selected features Naïve Bayes and Bayes net algorithms were used for classification. The classification results of Naïve Bayes algorithms and Bayes net algorithm for fault diagnosis of a hydraulic brake system were compared and the results were tabulated.