Indexed In Scopus
  Scopus ID: 21100926589

Fault Detection and Diagnosis of Rotor-ball Bearing System

Prince Shukla and Praveen Kumar Agarwal

Abstract

Every production machinery with a rotating element for transmitting power utilizes bearings to minimize frictional losses while supporting the rotating element and preventing the element’s motion in the direction of the applied load. This feature of the bearing makes it an essential element to watch for, as its failure may lead to low productivity of the production system and, if not taken proper care of, can even result in catastrophic failure of the production system. Consequently, in the present work, fault detection and diagnosis (FDD) of the rotor-ball bearing system is presented to diagnose the faults in the inner race, outer race, and balls of the bearing. As the vibration signals of the rotor-bearing system vary with healthy or different faulty conditions of ball bearing, these signals are used for FDD by using statistical feature extraction techniques to observe a recognizable pattern. These features’ values are now further used to train Artificial Neural Networks (ANN) (Feed-forward Backpropagation Neural Network; FBNN and Cascade Feed-forward Backpropagation Neural Network; CFBNN) to enable an automatic fault detection and diagnosis system. The main drawback of the Neural Network is its computational time requirement which has been overcome by using statistical feature extraction techniques while utilizing the adaptive nature of the Neural Network. The results show that Cascade Feedforward Backpropagation Neural Network gives better classification results and is more efficient than Feed-forward Backpropagation Neural Network.

Published on: May 24, 2023
doi: 10.17756/nwj.2023-s1-126
Citation: Shukla P, Agarwal PK. 2023. Fault Detection and Diagnosis of Rotor-ball Bearing System.NanoWorld J 9(S1): S653-S659.

| Download Citation (XML)

114 Downloads
Bitnami