Extraction of Key Environmental Factors and Construction of the Data-Driven Model for Atmospheric Aging of Polyurethane Coatings

Coating protection is an economical and effective means to prevent the corrosion of metal materials during service. However, organic coatings exposed to the atmospheric environment will also be subject to aging due to various environmental factors. In this paper, aiming at the problem that the mechanism of polyurethane coating atmospheric aging system is unclear and the aging law is difficult to analyze quantitatively, a framework for mining the influencing law of polyurethane coating atmospheric aging is proposed. Firstly, the Clustering Feature Elimination (CFE) algorithm is proposed to solve the problem that the key environmental factors are difficult to determine due to the complex mechanism of coating atmospheric aging. Secondly, based on Random Forest Regression (RFR), the atmospheric aging data-driven model of Polyurethane coating with key environmental factors as input is constructed. Finally, the influence rules of key environmental factors are mined using the model visualization method. The results show that the key environmental factors screened by the CFE algorithm are more consistent with the general law of coating atmospheric aging. The atmospheric aging model of polyurethane coating based on these key environmental factors has higher prediction accuracy. The coating aging law obtained by four-dimensional visualization can further reveal the mechanism of coating atmospheric aging.