A thermodynamics-consistent machine learning approach for ammonia-water thermal cycles

The integration of physics with data-driven models has emerged as a promising approach. However, the application of such hybrid approaches to ammonia-water thermal systems remains underexplored. We addressed this gap by developing a framework that integrates physical constraints through hard constra...

Full description

Saved in:
Bibliographic Details
Published in:Energy (Oxford) 2025-01, Vol.315, p.134443, Article 134443
Main Authors: Chen, Xin, Zhang, Lin, Huang, JiangBo, Jin, Lei, Song, YongShi, Zheng, XianHua, Zou, ZhiXiong
Format: Article
Language:English
Subjects:
Ammonia-water cycles
Constrained optimization
Data-driven modeling
Domain knowledge integration
Thermodynamic analysis
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The integration of physics with data-driven models has emerged as a promising approach. However, the application of such hybrid approaches to ammonia-water thermal systems remains underexplored. We addressed this gap by developing a framework that integrates physical constraints through hard constraint layers, soft loss penalties, and anomaly detection techniques. We validated this framework across three case studies including heat exchanger, absorption refrigeration, and Rankine cycle. To evaluate the performance, we introduced a novel Thermal Fitting Score (TFS) that combines the coefficient of determination, R2, and thermal inconsistency metrics. Our key contributions include (1) comprehensive exploratory data analysis for thermal cycle understanding, (2) thermal constraint formulation based on thermodynamic laws, and (3) constraint integration at architecture, training, and data levels. The constrained models achieve 100 % thermal law compliance with TFS improvements of 23 % and 63.5 % for heat exchanger and absorption refrigeration cases, respectively. This methodology advances the integration of thermal domain knowledge with data-driven approaches, ensuring both prediction accuracy and thermal consistency. •Appropriate data preprocessing and constraint formulation are conducted.•Soft-constrained, hard-constrained, and anomaly detection methods are integrated.•A new metric is proposed, named thermal fitting score (TFS).•The constrained models are compared against various baselines.•TFS is improved by 23 % and 63.5 % in case studies with limited training data.
ISSN:0360-5442
DOI:10.1016/j.energy.2025.134443