A Machine Learning Approach for Vibration Signal Based Fault Classification on Hydraulic Braking System through C4.5 Decision Tree Classifier and Logistic Model Tree Classifier 2020-28-0496

Car hydraulic brakes are important safety components for passengers and are thus the good condition of brakes are essential for braking. By using the vibrational signatures, the state of the brake components can be determined. In this proposed study, electronic condition monitoring is suggested as a possible solution to such issues by using a machine learning method with a piezo-electric transducer and a dynamic data acquisition system. Ford EcoSport setup was used to acquire the vibration signals for both good and bad braking conditions. The mathematical Descriptive statistical features from the vibration signals were obtained and the feature selection has been done with the C4.5 decision tree classifier. The appropriate number of features needed to classify a particular problem is not determined by a specific method. A thorough study is, therefore, necessary to find the right number of features. The fault analysis of the Ford EcoSport hydraulic braking system has been established through the use of the C4.5 decision tree classifier and logistic model tree (LMT) classifier.