Comparative analysis of ensemble, supervised, and deep learning regression algorithms for parametric modelling of solid-liquid fluidization

IF 5.5 3区工程技术 Q1 ENGINEERING, CHEMICAL

Journal of the Taiwan Institute of Chemical Engineers Pub Date : 2025-03-14 DOI:10.1016/j.jtice.2025.106053

Asif Afzal , Abdulrajak Buradi , Md. Tariqul Islam , Mohammad Asif , H. Fayaz , Sung Goon Park , Arunkumar Munimathan , Stéphane PA Bordas

{"title":"Comparative analysis of ensemble, supervised, and deep learning regression algorithms for parametric modelling of solid-liquid fluidization","authors":"Asif Afzal , Abdulrajak Buradi , Md. Tariqul Islam , Mohammad Asif , H. Fayaz , Sung Goon Park , Arunkumar Munimathan , Stéphane PA Bordas","doi":"10.1016/j.jtice.2025.106053","DOIUrl":null,"url":null,"abstract":"<div><h3>Background</h3><div>A comparative regression modelling of fluidization bed data parameters is performed in this work using different algorithms. Computational fluid dynamics (CFD) modelling of particle and fluid flow characters using two-fluid Eulerian-Eulerian model. RNG k-ε turbulence coupled with kinetic theory of granular flow was also combined. The developed numerical model is used for generating the fluidization related data of parameters like turbulent viscosity, turbulent dissipation rate, solid velocity, solid volume fraction, granular temperature, and turbulent kinetic energy.</div></div><div><h3>Methods</h3><div>Comparative modelling and performance analysis between ensemble learning, supervised learning, and neural networks is performed for the mentioned fluidized bed parameters. Ensemble Regression algorithms: Gradient boosting regressor (GBR), Voting regressor (VR), and Random-forest regressor (RFR), supervised learning algorithm - Decision tree (DT), and Deep Artificial neural network (ANN) models are used for the data mapping of fluidization parameters. Performance metrices are accessed in details to compare the modelling results or the algorithms in details for each fluidization parameter.</div></div><div><h3>Findings</h3><div>From the modelling of this data it is found that numerical data is highly non-linear. DT and RFR algorithms are the most accurate algorithms that predicted with >90 % of accuracy in each case. VT and GBR trained and tested with around 85 % accuracy in most cases but failed in prediction of granular temperature. ANN also sufficiently provided good accuracy while it also failed to predict granular temperature. Solid volume fraction, turbulent kinetic energy, turbulent viscosity, and turbulent dissipation rate were modelled perfectly with all the algorithms. Among all the parameters, turbulent viscosity during training and testing from each model is highly accurately modelled from each of the algorithm with prediction accuracy >90 %.</div></div>","PeriodicalId":381,"journal":{"name":"Journal of the Taiwan Institute of Chemical Engineers","volume":"171 ","pages":"Article 106053"},"PeriodicalIF":5.5000,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of the Taiwan Institute of Chemical Engineers","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1876107025001063","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, CHEMICAL","Score":null,"Total":0}

引用次数: 0

Abstract

Background

A comparative regression modelling of fluidization bed data parameters is performed in this work using different algorithms. Computational fluid dynamics (CFD) modelling of particle and fluid flow characters using two-fluid Eulerian-Eulerian model. RNG k-ε turbulence coupled with kinetic theory of granular flow was also combined. The developed numerical model is used for generating the fluidization related data of parameters like turbulent viscosity, turbulent dissipation rate, solid velocity, solid volume fraction, granular temperature, and turbulent kinetic energy.

Methods

Comparative modelling and performance analysis between ensemble learning, supervised learning, and neural networks is performed for the mentioned fluidized bed parameters. Ensemble Regression algorithms: Gradient boosting regressor (GBR), Voting regressor (VR), and Random-forest regressor (RFR), supervised learning algorithm - Decision tree (DT), and Deep Artificial neural network (ANN) models are used for the data mapping of fluidization parameters. Performance metrices are accessed in details to compare the modelling results or the algorithms in details for each fluidization parameter.

Findings

From the modelling of this data it is found that numerical data is highly non-linear. DT and RFR algorithms are the most accurate algorithms that predicted with >90 % of accuracy in each case. VT and GBR trained and tested with around 85 % accuracy in most cases but failed in prediction of granular temperature. ANN also sufficiently provided good accuracy while it also failed to predict granular temperature. Solid volume fraction, turbulent kinetic energy, turbulent viscosity, and turbulent dissipation rate were modelled perfectly with all the algorithms. Among all the parameters, turbulent viscosity during training and testing from each model is highly accurately modelled from each of the algorithm with prediction accuracy >90 %.

Abstract Image

查看原文本刊更多论文

固液流化参数化建模的集成、监督和深度学习回归算法的比较分析

本文采用不同的算法对流化床数据参数进行了比较回归建模。采用双流体欧拉-欧拉模型对颗粒和流体流动特性进行计算流体动力学（CFD）建模。RNG k-ε湍流与颗粒流动动力学理论相结合。所建立的数值模型用于生成湍流粘度、湍流耗散率、固体速度、固体体积分数、颗粒温度、湍流动能等参数的流化相关数据。方法对上述流化床参数进行集成学习、监督学习和神经网络的对比建模和性能分析。集成回归算法：梯度增强回归器（GBR），投票回归器（VR）和随机森林回归器（RFR），监督学习算法-决策树（DT）和深度人工神经网络（ANN）模型用于流化参数的数据映射。性能指标的详细访问，以比较建模结果或详细的算法为每个流化参数。从该数据的建模中发现，数值数据是高度非线性的。DT和RFR算法是最准确的算法，在每种情况下预测准确率为90%。VT和GBR在大多数情况下训练和测试的准确率约为85%，但在预测颗粒温度方面失败。人工神经网络也提供了足够好的精度，但它也不能预测颗粒温度。固体体积分数、湍流动能、湍流粘度和湍流耗散率均得到了较好的模拟。在所有参数中，每个模型的训练和测试过程中的湍流粘度都是由每个算法高度精确地建模的，预测精度为90%。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Journal of the Taiwan Institute of Chemical Engineers 工程技术-工程：化工

CiteScore

9.10

自引率

14.00%

发文量

362

审稿时长

35 days

期刊介绍： Journal of the Taiwan Institute of Chemical Engineers (formerly known as Journal of the Chinese Institute of Chemical Engineers) publishes original works, from fundamental principles to practical applications, in the broad field of chemical engineering with special focus on three aspects: Chemical and Biomolecular Science and Technology, Energy and Environmental Science and Technology, and Materials Science and Technology. Authors should choose for their manuscript an appropriate aspect section and a few related classifications when submitting to the journal online.