Intelligent fault inference for rotating flexible rotors using Bayesian belief network

  • Authors:
  • Bin Gang Xu

  • Affiliations:
  • Institute of Textiles and Clothing, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong

  • Venue:
  • Expert Systems with Applications: An International Journal
  • Year:
  • 2012

Quantified Score

Hi-index 12.05

Visualization

Abstract

Flexible rotor is a crucial mechanical component of a diverse range of rotating machineries and its condition monitoring and fault diagnosis are of particular importance to the modern industry. In this paper, Bayesian belief network (BBN) is applied to the fault inference for rotating flexible rotors with attempt to enhance the reasoning capacity under conditions of uncertainty. A generalized three-layer configuration of BBN for the fault inference of rotating machinery is developed by fully incorporating human experts' knowledge, machine faults and fault symptoms as well as machine running conditions. Compared with the Naive diagnosis network, the proposed topological structure of causalities takes account of more practical and complete diagnostic information in fault diagnosis. The network tallies well with the practical thinking of field experts in the whole processes of machine fault diagnosis. The applications of the proposed BBN network in the uncertainty inference of rotating flexible rotors show good agreements with our knowledge and practical experience of diagnosis.