海洋环境下3C钢腐蚀速度多元非线性回归模型的建立及验证

Establishment and verification of multivariate nonlinear regression model for corrosion rate of 3C steel in marine environment

通过分析海水环境因素对钢材腐蚀速度的影响,采用多元非线性分析方法建立了计算3C钢腐蚀速度的多元非线性数学模型,并将其计算结果与线性回归模型的计算结果及实验结果进行了对比。结果表明:在考虑各自变量相互作用的情况下,本文建立的多元非线性模型的平均相对误差为8.249 %,均方误差为1.607,其精度高于未考虑各自变量相互作用的非线性模型精度;而线性模型的平均相对误差为12.406 %,均方误差为4.169,其精度低于非线性模型;采用三次多项式来描述腐蚀速度与温度和盐度之间的关系是合理的。本文研究结果可为海洋环境下3C钢腐蚀速度的预测提供有益借鉴。

By analyzing the influence of seawater environmental factors on steel corrosion rate, a multivariate non-linear mathematical model for calculating corrosion rate of 3C steel was established by using multivariate non-linear analysis method, and the calculated results were compared with those of linear regression model and experimental results. The results show that the average relative error and mean square error of the multivariate nonlinear model in this paper are 8.249 % and 1.607 respectively when considering the interaction of the variables, and the precision is higher than that of the nonlinear model without considering the interaction among them. The average relative error and mean square error of the linear model are 12.406 % and 4.169 respectively, and the precision of the linear model is lower than that of the nonlinear model. It is reasonable to use cubic polynomial to describe the relationship of corrosion rate with temperature and salinity and corrosion rate. The results of this study can provide useful reference for predicting corrosion rate of 3C steel in marine environment.