Vibration and Support Vector Machine-Based Fault Diagnosis of Bevel Gearbox

Abstract

There is a rapid increase in the number of problems associated with initial fault detection in gearboxes. A minor defect in a gear tooth raises the gearbox's vibrations, resulting in catastrophic failure

therefore, it is essential to detect and correct the defect as soon as feasible. The defects may be diagnosed by using machine learning (ML) algorithms. This chapter presents an approach to the diagnosis of the fault in a bevel gearbox by using a support vector machine (SVM), which is a supervised ML algorithm. The present method requires the collection of vibration signals from the bevel gearbox and the extraction of time domain statistical features, such as mean, standard deviation, kurtosis, and skewness. Then, these features are used in the SVM classifier to distinguish between normal and malfunctioning transmission conditions. The SVM classification accuracy analysis is carried out by using different kernel functions. The RBF kernel function has been observed to be a more effective kernel than the other two linear and exponential kernel functions. The results demonstrated that the proposed method outperformed conventional techniques by detecting defects with greater precision. © 2025 selection and editorial matter, Deepam Goyal, Ankit Sharma and Mohamad Abou Houran

individual chapters, the contributors.