A Methodology for Automated Pellet size Distribution in a Pellet Mill

David James, Mauro Innocente, John Cherry, David Carswell, Marc Holmes, Steve Brown, Nick Lavery

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

In this work the feasibility of pellet size distribution using images captured from the operator’s sight glass of the moving pellet bed using a standard digital camera is investigated. The pellet size distribution is determined using a bespoke circle detection program which is able to fit circles to pellet edges, neural networks are used to match predicted distributions to actual distributions and discrete element modelling is used to create a large quantity of rendered images with known distributions for network training. The results show that the circle detection algorithm is able to find a high percentage of pellets in an image and find their sizes. Neural networks are then able to use the predicted pellet size distributions to give a good prediction of the actual distributions and are able to account for segregation effects in the pellet beds. This demonstrates that an online pellet sizing scheme is feasible in nickel pellet production.
Original languageEnglish
Title of host publicationKES Transactions on Sustainable Design and Manufacturing
EditorsR. Setchi, R.J. Howlett, M. Naim, J. Sienz
Place of PublicationEngland, UK
PublisherFuture Technology Press
Pages732-743
Number of pages12
Volume1
ISBN (Print)978-0-9561516-5-0, 978-0-9561516-6-7, 978-0-9561516-4-3
Publication statusPublished - 30 Apr 2014
Externally publishedYes
EventInternational Conference on Sustainable Design and Manufacturing - Cardiff, United Kingdom
Duration: 28 Apr 201430 Apr 2014

Conference

ConferenceInternational Conference on Sustainable Design and Manufacturing
CountryUnited Kingdom
CityCardiff
Period28/04/1430/04/14

Fingerprint

Neural networks
Digital cameras
Nickel
Glass

Cite this

James, D., Innocente, M., Cherry, J., Carswell, D., Holmes, M., Brown, S., & Lavery, N. (2014). A Methodology for Automated Pellet size Distribution in a Pellet Mill. In R. Setchi, R. J. Howlett, M. Naim, & J. Sienz (Eds.), KES Transactions on Sustainable Design and Manufacturing (Vol. 1, pp. 732-743). [sdm14-108] England, UK: Future Technology Press.

A Methodology for Automated Pellet size Distribution in a Pellet Mill. / James, David; Innocente, Mauro; Cherry, John; Carswell, David; Holmes, Marc; Brown, Steve; Lavery, Nick.

KES Transactions on Sustainable Design and Manufacturing. ed. / R. Setchi; R.J. Howlett; M. Naim; J. Sienz. Vol. 1 England, UK : Future Technology Press, 2014. p. 732-743 sdm14-108.

Research output: Chapter in Book/Report/Conference proceedingChapter

James, D, Innocente, M, Cherry, J, Carswell, D, Holmes, M, Brown, S & Lavery, N 2014, A Methodology for Automated Pellet size Distribution in a Pellet Mill. in R Setchi, RJ Howlett, M Naim & J Sienz (eds), KES Transactions on Sustainable Design and Manufacturing. vol. 1, sdm14-108, Future Technology Press, England, UK, pp. 732-743, International Conference on Sustainable Design and Manufacturing, Cardiff, United Kingdom, 28/04/14.
James D, Innocente M, Cherry J, Carswell D, Holmes M, Brown S et al. A Methodology for Automated Pellet size Distribution in a Pellet Mill. In Setchi R, Howlett RJ, Naim M, Sienz J, editors, KES Transactions on Sustainable Design and Manufacturing. Vol. 1. England, UK: Future Technology Press. 2014. p. 732-743. sdm14-108
James, David ; Innocente, Mauro ; Cherry, John ; Carswell, David ; Holmes, Marc ; Brown, Steve ; Lavery, Nick. / A Methodology for Automated Pellet size Distribution in a Pellet Mill. KES Transactions on Sustainable Design and Manufacturing. editor / R. Setchi ; R.J. Howlett ; M. Naim ; J. Sienz. Vol. 1 England, UK : Future Technology Press, 2014. pp. 732-743
@inbook{ce25edb1e83a4771ad45146229a712da,
title = "A Methodology for Automated Pellet size Distribution in a Pellet Mill",
abstract = "In this work the feasibility of pellet size distribution using images captured from the operator’s sight glass of the moving pellet bed using a standard digital camera is investigated. The pellet size distribution is determined using a bespoke circle detection program which is able to fit circles to pellet edges, neural networks are used to match predicted distributions to actual distributions and discrete element modelling is used to create a large quantity of rendered images with known distributions for network training. The results show that the circle detection algorithm is able to find a high percentage of pellets in an image and find their sizes. Neural networks are then able to use the predicted pellet size distributions to give a good prediction of the actual distributions and are able to account for segregation effects in the pellet beds. This demonstrates that an online pellet sizing scheme is feasible in nickel pellet production.",
author = "David James and Mauro Innocente and John Cherry and David Carswell and Marc Holmes and Steve Brown and Nick Lavery",
year = "2014",
month = "4",
day = "30",
language = "English",
isbn = "978-0-9561516-5-0",
volume = "1",
pages = "732--743",
editor = "R. Setchi and R.J. Howlett and M. Naim and J. Sienz",
booktitle = "KES Transactions on Sustainable Design and Manufacturing",
publisher = "Future Technology Press",

}

TY - CHAP

T1 - A Methodology for Automated Pellet size Distribution in a Pellet Mill

AU - James, David

AU - Innocente, Mauro

AU - Cherry, John

AU - Carswell, David

AU - Holmes, Marc

AU - Brown, Steve

AU - Lavery, Nick

PY - 2014/4/30

Y1 - 2014/4/30

N2 - In this work the feasibility of pellet size distribution using images captured from the operator’s sight glass of the moving pellet bed using a standard digital camera is investigated. The pellet size distribution is determined using a bespoke circle detection program which is able to fit circles to pellet edges, neural networks are used to match predicted distributions to actual distributions and discrete element modelling is used to create a large quantity of rendered images with known distributions for network training. The results show that the circle detection algorithm is able to find a high percentage of pellets in an image and find their sizes. Neural networks are then able to use the predicted pellet size distributions to give a good prediction of the actual distributions and are able to account for segregation effects in the pellet beds. This demonstrates that an online pellet sizing scheme is feasible in nickel pellet production.

AB - In this work the feasibility of pellet size distribution using images captured from the operator’s sight glass of the moving pellet bed using a standard digital camera is investigated. The pellet size distribution is determined using a bespoke circle detection program which is able to fit circles to pellet edges, neural networks are used to match predicted distributions to actual distributions and discrete element modelling is used to create a large quantity of rendered images with known distributions for network training. The results show that the circle detection algorithm is able to find a high percentage of pellets in an image and find their sizes. Neural networks are then able to use the predicted pellet size distributions to give a good prediction of the actual distributions and are able to account for segregation effects in the pellet beds. This demonstrates that an online pellet sizing scheme is feasible in nickel pellet production.

M3 - Chapter

SN - 978-0-9561516-5-0

SN - 978-0-9561516-6-7

SN - 978-0-9561516-4-3

VL - 1

SP - 732

EP - 743

BT - KES Transactions on Sustainable Design and Manufacturing

A2 - Setchi, R.

A2 - Howlett, R.J.

A2 - Naim, M.

A2 - Sienz, J.

PB - Future Technology Press

CY - England, UK

ER -