Advances in the application of machine learning to boiling heat transfer: A review

Boiling heat transfer, one of the most common and effective heat dissipation methods, is prevalent in industries and crucial for cooling electronic components such as chips. The key to boiling heat transfer research lies in enhancing improve its heat transfer performance, which is typically characterized by complex physical phenomena. Moreover, how to accurately predict the heat transfer process is still an important problem to be solved. Boiling heat transfer is generally associated with multiple parameters that are not accurately predicted by the usual mathematical form of empirical correlations. Therefore, this paper mainly reviews the applications of machine learning in predicting boiling heat transfer in recent years and provides a brief introduction to the main machine learning algorithms currently in use. The paper discusses the application of machine learning for predicting important parameters such as heat flux, heat transfer coefficient and critical heat flux, and examines the limitations of machine learning in boiling heat transfer researches.