A physically interpretable machine-learning method in efficiently and explicitly exploring predictive thermodynamic models for mixture working fluids

doi:10.1016/j.energy.2025.139577

A physically interpretable machine-learning method in efficiently and explicitly exploring predictive thermodynamic models for mixture working fluids

2025-12-11

0

PRE

AI

DOI：10.1016/j.energy.2025.139577

Original

Original request for help

Save

Abstract

En 中文

• Refine physics-embedded symbolic regression framework for exploring explicit models. • Develop a visual statistics and screening method for algebraic elements. • Establish a universal and reliable prediction model for mixture speed of sound. • Verify model applicability on various working fluids over wide thermodynamic ranges. • Reveal the mathematical stability and physical significance of the model.

A physically interpretable machine-learning method in efficiently and explicitly exploring predictive thermodynamic models for mixture working fluids

Abstract

Journal

Organization