Machine learning-based approach for predicting flow boiling heat transfer coefficient at high saturation temperatures

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F46747885%3A24210%2F25%3A00012803" target="_blank" >RIV/46747885:24210/25:00012803 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.sciencedirect.com/science/article/pii/S0735193324013009" target="_blank" >https://www.sciencedirect.com/science/article/pii/S0735193324013009</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.icheatmasstransfer.2024.108538" target="_blank" >10.1016/j.icheatmasstransfer.2024.108538</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Machine learning-based approach for predicting flow boiling heat transfer coefficient at high saturation temperatures

Popis výsledku v původním jazyce

This paper explores the prediction performance of nine (9) machine learning (ML) models at high saturation temperatures for which most empirical models have been reported to fail. Over thousand five hundred data points were carefully compiled from seven different authors utilizing three different working fluids that have been applied or recommended to work well in high temperature applications. Both dimensional and dimensionless features of the dataset were evaluated and compared. Results indicate that machine learning models offer enhanced accuracy compared to empirical models. From the nine machine learning models evaluated, for the case of dimensional features, Gradient Boosting, XGBoost, K-Nearest Neighbor, and Random Forest were the best performing models with Mean Absolute Errors (MAEs) less than 10 % and R-square values over 95 %. In the case of dimensionless features, Gradient Boosting, XGBoost, Random Forest, and Extra Tree were the best-performing models with Mean Absolute Errors (MAEs) less than 10 % and R-square values over 95 %. Overall, XGBoost, Gradient Boosting, and Random Forest were the models that remained resolute in their performance when the data was transformed from dimensional to dimensionless features. Feature importance was also performed to rank the features on how they contributed to the models‘ prediction.

Název v anglickém jazyce

Machine learning-based approach for predicting flow boiling heat transfer coefficient at high saturation temperatures

Popis výsledku anglicky

This paper explores the prediction performance of nine (9) machine learning (ML) models at high saturation temperatures for which most empirical models have been reported to fail. Over thousand five hundred data points were carefully compiled from seven different authors utilizing three different working fluids that have been applied or recommended to work well in high temperature applications. Both dimensional and dimensionless features of the dataset were evaluated and compared. Results indicate that machine learning models offer enhanced accuracy compared to empirical models. From the nine machine learning models evaluated, for the case of dimensional features, Gradient Boosting, XGBoost, K-Nearest Neighbor, and Random Forest were the best performing models with Mean Absolute Errors (MAEs) less than 10 % and R-square values over 95 %. In the case of dimensionless features, Gradient Boosting, XGBoost, Random Forest, and Extra Tree were the best-performing models with Mean Absolute Errors (MAEs) less than 10 % and R-square values over 95 %. Overall, XGBoost, Gradient Boosting, and Random Forest were the models that remained resolute in their performance when the data was transformed from dimensional to dimensionless features. Feature importance was also performed to rank the features on how they contributed to the models‘ prediction.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

20303 - Thermodynamics

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach<br>I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

International Communications in Heat and Mass Transfer

ISSN

0735-1933

e-ISSN

—

Svazek periodika

161

Číslo periodika v rámci svazku

31 December 2024

Stát vydavatele periodika

NL - Nizozemsko

Počet stran výsledku

18

Strana od-do

—

Kód UT WoS článku

001421918300001

EID výsledku v databázi Scopus

2-s2.0-85213524282