Enhancing random forest model prediction of gas holdup in internal draft airlift loop contactors with genetic algorithms tuning and interpretability DOI Creative Commons
Hardik Prabhu,

C. Ravishankar,

A. Ganesan

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 18, 2025

Gas holdup in an Internal Loop Airlift Contactor (ILAC) is one of the primary hydrodynamic features that governs performance contactor. Prediction gas assumes importance design airlift The key challenge prediction due to difficulty obtaining reliable phenomenological models. improved accuracies can provide better contactor and equipment performance. current work focuses on use data driven models for holdup, as existing empirical correlations were found be inadequate. In this work, 324 points from reported works internal draft are consolidated. input part examples set consists ten which related geometry, fluid properties operating conditions. This study improves predictive accuracy output variable, ILACs using a Random Forest (RF) model optimized via Genetic Algorithms (GA). SHapley Additive exPlanations (SHAP) based interpretability reveals contribution features. tuned achieved Coefficient Determination (R²) score 0.9542 Mean Absolute Error (MAE) 0.0059, surpassing traditional parameter sets providing insights into control.

Language: Английский

Enhancing random forest model prediction of gas holdup in internal draft airlift loop contactors with genetic algorithms tuning and interpretability DOI Creative Commons
Hardik Prabhu,

C. Ravishankar,

A. Ganesan

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 18, 2025

Gas holdup in an Internal Loop Airlift Contactor (ILAC) is one of the primary hydrodynamic features that governs performance contactor. Prediction gas assumes importance design airlift The key challenge prediction due to difficulty obtaining reliable phenomenological models. improved accuracies can provide better contactor and equipment performance. current work focuses on use data driven models for holdup, as existing empirical correlations were found be inadequate. In this work, 324 points from reported works internal draft are consolidated. input part examples set consists ten which related geometry, fluid properties operating conditions. This study improves predictive accuracy output variable, ILACs using a Random Forest (RF) model optimized via Genetic Algorithms (GA). SHapley Additive exPlanations (SHAP) based interpretability reveals contribution features. tuned achieved Coefficient Determination (R²) score 0.9542 Mean Absolute Error (MAE) 0.0059, surpassing traditional parameter sets providing insights into control.

Language: Английский

Citations

0