Machine Learning-Based Prediction of Tribological Properties of Epoxy Composite Coating

Machine learning, being convenient and nondestructive, is beneficial for evaluating the tribological properties of coatings. Here, six machine learning algorithms, using a sericite/epoxy composite coating (SEC) as an example, were employed to assess the impact of filler content (10, 15, 20, 25, and 30 wt%) and mesh size on the tribological properties of epoxy composite coatings under different loads. The results showed that the gradient boosting regression model had superior accuracy and stability compared to the other regression models, achieving friction coefficient and wear rate prediction accuracies of 93.7% and 85.7%, respectively. This model outperformed others, including decision trees, extreme gradient boosting, and Gaussian process regression. Feature importance showed that the content of sericite had the most significant influence on the tribological properties. This work provides valuable guidance for the engineering application of this material.