Artificial Neural Networks (ANNs)

Asoke Nandi, Hosameldin Ahmed

Research output: Chapter in Book/Published conference outputChapter

Abstract

Artificial neural network (ANN) often consists of a series of algorithms that work together to recognise underlying relationships in a set of data. This chapter introduces some widely used ANN algorithms that have been used for machine fault diagnosis using vibration signals. It presents essential concepts of ANNs; then describes three different types of ANN (i.e. multilayer perceptron, radial basis function network, and Kohonen network) that can be used for fault classification. In addition, the chapter describes the applications of these methods and several other types of ANN-based methods in machine fault diagnosis. A considerable amount of literature has been published on the application of ANNs and variants in machine fault diagnosis. Most of these studies introduced many preprocessing techniques that include normalisation, feature selection, transformation, and feature extraction. The produced data of the preprocessing step represent the final training set that is used as input to ANNs.
Original languageEnglish
Title of host publicationCondition Monitoring with Vibration Signals
Chapter12
Pages239-258
DOIs
Publication statusPublished - 6 Dec 2019

Fingerprint

Dive into the research topics of 'Artificial Neural Networks (ANNs)'. Together they form a unique fingerprint.

Cite this