Research on AI-Based Optimization Design of Energy Heat Transfer Systems
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Using artificial intelligence-based analytical methods, this study conducted structured modeling and optimization research on operational data from energy heat transfer systems. By collecting 4,956 multi-state operational samples, classification recognition and energy efficiency regression models were constructed, enabling intelligent identification of system states and precise prediction of the coefficient of performance (COP). A multi-objective evolutionary algorithm was employed for comprehensive optimization of heat exchange power, pressure differential, and energy efficiency, yielding a Pareto front solution set. Results indicate a 96.2% accuracy rate for normal state identification. For COP prediction, the LSTM model achieved RMSE values of 0.091 and 0.138 under normal and low-efficiency states respectively, with R2reaching up to 0.948. A three-objective optimization model combined with the NSGA-II algorithm generated a Pareto optimal solution set satisfying operational constraints. Results demonstrate the system’s strong performance in both accuracy and adaptability.
Fields of Study & Subjects
Building Energy and Comfort Optimization
Refrigeration and Air Conditioning Technologies
Heat Transfer and Optimization
Publication details
- Journal / source
- IEEE International Conference on Power, Electronics and Green Energy (ICPEGE)
- Pages
- 1538–1542
- Publication date
- 28 January 2026
- Language
- EN