Vibration Based Wheel Condition Monitoring System Using Machine Learning Approach

Abstract

Wheel condition monitoring system (WCMS) is an emerging electronic safety system newlinepresent in most of the new generation vehicles. It is intended to warn the driver newlinewhen wheels are unbalanced or inflation pressure in a tyre goes below a certain recommended newlinevalue. In India only limited vehicles having this critical safety system installed newlinebecause it is highly expensive. The existing WCMS monitor only the tyre pressure newlineusing direct or indirect methods. In recent days the indirect WCMS using vibration signals newlineand machine learning techniques are gaining momentum. The proposed study was newlinecarried out using vibration signal data acquired from a low cost accelerometer placed newlineon the wheel hub of both air filled and nitrogen filled tyres. The features required newlinewere taken out from the acquired data with the help of statistical, histogram and auto newlineregressive-moving average (ARMA) feature extraction techniques. Features selection newlinewas carried out using J48 decision tree algorithm. For feature classification seven classifier newlinefamilies like tree based, meta, rules, bayes, functions and lazy classifiers were newlineused (ie.,) totally 39 classifiers have been used for WCMS. The best feature-classifier newlinepair for WCMS, for both air-filled and nitrogen-filled tyres, is proposed by this study. newlineThe present study goes one step forward to include the wheel balancing in the WCMS, newlinethus increasing the scope of WCMS. newline