Erosion modelling using Bayesian regulated artificial neural networks

Sean Danaher; Psantu Datta; Philip Hackney; I. Waddle

doi:10.1016/j.wear.2003.08.006

Erosion modelling using Bayesian regulated artificial neural networks

Sean Danaher, Psantu Datta, Philip Hackney, I. Waddle

Research output: Contribution to journal › Article › peer-review

37 Citations (Scopus)

Abstract

Modelling of the high temperature erosion behaviour of Ni-base alloys using artificial neural networks (ANNs) is presented. Two scenarios have been used: (i) a simple equation-based model and (ii) a comprehensive dataset looking at erosion as a function of particle size, velocity, impact angle and temperature. Common problems associated with ANNs are discussed within the context of erosion modelling. It has been found that the use of multilayer perceptron artificial neural networks for modelling erosion gave unreliable results when trained with traditional algorithms. The more recent Bayesian regularisation algorithm however has proved very successful, yielding both high Pearsonian correlation coefficients (r>0.95) and accuracies averaging better than 90%.

Original language	English
Pages (from-to)	879-888
Journal	Wear
Volume	256
Issue number	9-10
DOIs	https://doi.org/10.1016/j.wear.2003.08.006
Publication status	Published - Apr 2004

Keywords

Neural networks (Computer science)

Access to Document

10.1016/j.wear.2003.08.006

Cite this

@article{b05b3c0865824ecf9ea714d98ec6123e,

title = "Erosion modelling using Bayesian regulated artificial neural networks",

abstract = "Modelling of the high temperature erosion behaviour of Ni-base alloys using artificial neural networks (ANNs) is presented. Two scenarios have been used: (i) a simple equation-based model and (ii) a comprehensive dataset looking at erosion as a function of particle size, velocity, impact angle and temperature. Common problems associated with ANNs are discussed within the context of erosion modelling. It has been found that the use of multilayer perceptron artificial neural networks for modelling erosion gave unreliable results when trained with traditional algorithms. The more recent Bayesian regularisation algorithm however has proved very successful, yielding both high Pearsonian correlation coefficients (r>0.95) and accuracies averaging better than 90\%.",

keywords = "Neural networks (Computer science)",

author = "Sean Danaher and Psantu Datta and Philip Hackney and I. Waddle",

note = "A collaborative work between Energy Systems, Materials \& Manufacturing research group and Northumbria Communication Research Laboratory, using the neural network for the first time to predict the behaviour of erosion. A good example of research collaboration between the two research groups illustrating true interdisciplinary research in the General Engineering context.",

year = "2004",

month = apr,

doi = "10.1016/j.wear.2003.08.006",

language = "English",

volume = "256",

pages = "879--888",

journal = "Wear",

issn = "0043-1648",

publisher = "Elsevier B.V.",

number = "9-10",

}

TY - JOUR

T1 - Erosion modelling using Bayesian regulated artificial neural networks

AU - Danaher, Sean

AU - Datta, Psantu

AU - Hackney, Philip

AU - Waddle, I.

N1 - A collaborative work between Energy Systems, Materials & Manufacturing research group and Northumbria Communication Research Laboratory, using the neural network for the first time to predict the behaviour of erosion. A good example of research collaboration between the two research groups illustrating true interdisciplinary research in the General Engineering context.

PY - 2004/4

Y1 - 2004/4

N2 - Modelling of the high temperature erosion behaviour of Ni-base alloys using artificial neural networks (ANNs) is presented. Two scenarios have been used: (i) a simple equation-based model and (ii) a comprehensive dataset looking at erosion as a function of particle size, velocity, impact angle and temperature. Common problems associated with ANNs are discussed within the context of erosion modelling. It has been found that the use of multilayer perceptron artificial neural networks for modelling erosion gave unreliable results when trained with traditional algorithms. The more recent Bayesian regularisation algorithm however has proved very successful, yielding both high Pearsonian correlation coefficients (r>0.95) and accuracies averaging better than 90%.

AB - Modelling of the high temperature erosion behaviour of Ni-base alloys using artificial neural networks (ANNs) is presented. Two scenarios have been used: (i) a simple equation-based model and (ii) a comprehensive dataset looking at erosion as a function of particle size, velocity, impact angle and temperature. Common problems associated with ANNs are discussed within the context of erosion modelling. It has been found that the use of multilayer perceptron artificial neural networks for modelling erosion gave unreliable results when trained with traditional algorithms. The more recent Bayesian regularisation algorithm however has proved very successful, yielding both high Pearsonian correlation coefficients (r>0.95) and accuracies averaging better than 90%.

KW - Neural networks (Computer science)

UR - https://www.scopus.com/pages/publications/2342627945

U2 - 10.1016/j.wear.2003.08.006

DO - 10.1016/j.wear.2003.08.006

M3 - Article

SN - 0043-1648

VL - 256

SP - 879

EP - 888

JO - Wear

JF - Wear

IS - 9-10

ER -

Erosion modelling using Bayesian regulated artificial neural networks

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this