Investigations on the effect of oil quality on gearboxes using neural network predictors

Abstract

Purpose – The purpose of this paper was to perform an experimental investigation to analyze vibration and noise of unloaded gearbox with different oil quality. All motor-driven machinery used in the modern world can develop faults. The maintenance plans include analyzing the external relevant information of critical components, in order to evaluate its internal state. From the beginning of the twentieth century, different technologies have been used to process signals of dynamical systems. Design/methodology/approach – A proposed neural network (NN) is also employed to predict vibration parameters of the experimental test rig. Moreover, four types of oils are used for gearbox to predict reliable oil. Vibration signals extracted from rotating parts of machineries carry lot many information within them about the condition of the operating machine. Further processing of these raw vibration signatures measured at a convenient location of the machine unravels the condition of the component or the assembly under study. The experimental stand for testing an unloaded gearbox is composed by actuated direct current (DC) driving system. Findings – This paper deals with the effectiveness of wavelet-based features for fault diagnosis of a gearbox using two types of artificial neural networks (ANNs) and stress analyzed with computer-based software ANNs. The results improved that the proposed NN has superior performance to adapt experimental results. Practical implications – This paper is one such attempt to apply machine learning methods like ANN. This work deals with extraction of wavelet features from the vibration data of a gearbox system and classification of gear faults using ANNs. Originality/value – These kind of NN-based approaches are novel approaches to predict real-time vibration and acceleration parameters of unloaded gearbox with five types of oils. Also, the investigation contains new information about studied process, containing elements of novelty.