A Comprehensive Approach for Detecting Brake Pad Defects Using Histogram and Wavelet Features with Nested Dichotomy...

Area of Science:

• Mechanical Engineering
• Automotive Engineering
• Signal Processing

Background:

• The hydraulic brake system is a critical safety component in commercial vehicles.
• Continuous monitoring is essential to ensure reliable brake system performance.
• Vibration analysis offers a promising non-invasive method for condition monitoring.

Purpose of the Study:

• To develop and evaluate a method for monitoring the hydraulic brake system using vibration signals.
• To accurately categorize brake conditions (good vs. defective) based on extracted features.
• To identify the most effective classification algorithm for this application.

Main Methods:

• Experimental capture of vibration signals from commercial vehicle brake pad assemblies under normal and defective conditions.
• Extraction of relevant features, including histograms and wavelet features.
• Classification of extracted features using Nested dichotomy family algorithms.

Main Results:

• Wavelet features combined with the class-balanced nested dichotomy algorithm achieved the highest accuracy.
• A maximum classification accuracy of 99.45% was obtained.
• The proposed method effectively distinguishes between healthy and defective brake systems.

Conclusions:

• Vibration signal analysis, particularly using wavelet features and advanced classification, provides a highly accurate method for hydraulic brake system monitoring.
• This approach ensures the reliability and safety of commercial vehicle braking systems.
• Implementation of this monitoring technique enhances overall vehicle safety and operational efficiency.