Machine Learning-Based Optimization Algorithm for Distributed Geometric Parameters of Heat Exchangers
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
The design of geometric parameters in heat exchangers exerts a notable influence on their heat transfer capabilities and fluid dynamic behaviors. Conventional optimization approaches usually depend on empirical formulas or numerical simulation methods, which are plagued by excessive computational expenses and low operational efficiency. To tackle this issue, this research puts forward a machine learning-driven optimization algorithm tailored for the distributed geometric parameters of heat exchangers. First, computational fluid dynamics (CFD) simulation technology is used to produce a large volume of sample data, which includes both the geometric parameters of heat exchangers and their corresponding performance metrics. This sample data is then used to build a comprehensive dataset. Next, machine learning tools—such as neural networks and random forests—are applied to develop prediction models that establish relationships between the geometric parameters of heat exchangers and their performance indexes. Finally, the study integrates intelligent optimization algorithms (e.g., genetic algorithms and particle swarm optimization) to realize the multi-objective optimization design of heat exchanger geometric parameters. The research findings indicate that this proposed method can remarkably boost optimization efficiency. Moreover, the optimized heat exchanger design not only enhances heat transfer performance but also effectively controls flow resistance. This provides a novel technical route for heat exchanger design, endowing the research with substantial theoretical significance and practical value in engineering applications.
Fields of Study & Subjects
Heat Transfer and Optimization
Advanced Multi-Objective Optimization Algorithms
Heat Transfer and Numerical Methods
Publication details
- Journal / source
- IEEE International Conference on Power, Electronics and Green Energy (ICPEGE)
- Pages
- 278–283
- Publication date
- 28 January 2026
- Language
- EN