Pressure drops of fresh cemented paste backfills through coupled test loop experiments and machine learning techniques

Chongchong Qi, Qiusong Chen, Xiangjian Dong, Qinli Zhang, Zaher Mundher Yaseen

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

In this paper, test loop experiments and machine learning techniques were combined to investigate pressure drops of fresh cemented paste backfill (CPB) mixes. The influence of tailings characteristics on CPB pressure drops was studied and Extensive test loop experiments were performed. The complex mapping from tailings characteristics, cement-tailings ratio, inlet velocity and solids content to pressure drop was successfully learned by decision tree regression (DTR) models. In addition, the influence of training set size and maximum tree depth on DTR performance was investigated. The relative importance of predictor variables was discussed and the visualisation for a representative DTR model was provided. Finally, the current research applied two ensemble techniques, namely random forest and gradient boosting regression tree, to increase the predictive performance of the DTR models. The study found that all ensemble techniques outperformed DTR in the pressure drop prediction of CPB.

Original languageEnglish
JournalPowder Technology
DOIs
Publication statusAccepted/In press - 22 Nov 2019

Fingerprint

Ointments
Decision trees
Pressure drop
Learning systems
Tailings
Experiments
Cements
Visualization

Cite this

@article{bffeb7b41c2244868aa3bfe4f9cdda52,
title = "Pressure drops of fresh cemented paste backfills through coupled test loop experiments and machine learning techniques",
abstract = "In this paper, test loop experiments and machine learning techniques were combined to investigate pressure drops of fresh cemented paste backfill (CPB) mixes. The influence of tailings characteristics on CPB pressure drops was studied and Extensive test loop experiments were performed. The complex mapping from tailings characteristics, cement-tailings ratio, inlet velocity and solids content to pressure drop was successfully learned by decision tree regression (DTR) models. In addition, the influence of training set size and maximum tree depth on DTR performance was investigated. The relative importance of predictor variables was discussed and the visualisation for a representative DTR model was provided. Finally, the current research applied two ensemble techniques, namely random forest and gradient boosting regression tree, to increase the predictive performance of the DTR models. The study found that all ensemble techniques outperformed DTR in the pressure drop prediction of CPB.",
keywords = "Cemented paste backfill, Decision tree regression, Ensemble technique, Machine learning, Pressure drop, Test loop",
author = "Chongchong Qi and Qiusong Chen and Xiangjian Dong and Qinli Zhang and Yaseen, {Zaher Mundher}",
year = "2019",
month = "11",
day = "22",
doi = "10.1016/j.powtec.2019.11.046",
language = "English",
journal = "Powder Technology",
issn = "0032-5910",
publisher = "Pergamon",

}

Pressure drops of fresh cemented paste backfills through coupled test loop experiments and machine learning techniques. / Qi, Chongchong; Chen, Qiusong; Dong, Xiangjian; Zhang, Qinli; Yaseen, Zaher Mundher.

In: Powder Technology, 22.11.2019.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Pressure drops of fresh cemented paste backfills through coupled test loop experiments and machine learning techniques

AU - Qi, Chongchong

AU - Chen, Qiusong

AU - Dong, Xiangjian

AU - Zhang, Qinli

AU - Yaseen, Zaher Mundher

PY - 2019/11/22

Y1 - 2019/11/22

N2 - In this paper, test loop experiments and machine learning techniques were combined to investigate pressure drops of fresh cemented paste backfill (CPB) mixes. The influence of tailings characteristics on CPB pressure drops was studied and Extensive test loop experiments were performed. The complex mapping from tailings characteristics, cement-tailings ratio, inlet velocity and solids content to pressure drop was successfully learned by decision tree regression (DTR) models. In addition, the influence of training set size and maximum tree depth on DTR performance was investigated. The relative importance of predictor variables was discussed and the visualisation for a representative DTR model was provided. Finally, the current research applied two ensemble techniques, namely random forest and gradient boosting regression tree, to increase the predictive performance of the DTR models. The study found that all ensemble techniques outperformed DTR in the pressure drop prediction of CPB.

AB - In this paper, test loop experiments and machine learning techniques were combined to investigate pressure drops of fresh cemented paste backfill (CPB) mixes. The influence of tailings characteristics on CPB pressure drops was studied and Extensive test loop experiments were performed. The complex mapping from tailings characteristics, cement-tailings ratio, inlet velocity and solids content to pressure drop was successfully learned by decision tree regression (DTR) models. In addition, the influence of training set size and maximum tree depth on DTR performance was investigated. The relative importance of predictor variables was discussed and the visualisation for a representative DTR model was provided. Finally, the current research applied two ensemble techniques, namely random forest and gradient boosting regression tree, to increase the predictive performance of the DTR models. The study found that all ensemble techniques outperformed DTR in the pressure drop prediction of CPB.

KW - Cemented paste backfill

KW - Decision tree regression

KW - Ensemble technique

KW - Machine learning

KW - Pressure drop

KW - Test loop

UR - http://www.scopus.com/inward/record.url?scp=85076251609&partnerID=8YFLogxK

U2 - 10.1016/j.powtec.2019.11.046

DO - 10.1016/j.powtec.2019.11.046

M3 - Article

JO - Powder Technology

JF - Powder Technology

SN - 0032-5910

ER -