经验参数预测高熵合金屈服强度的多元线性回归模型

Multiple Linear Regression Model for Predicting Yield Strength of HEA by EmpiricalParameters

提出了一种基于经验参数来预测高熵合金(HEA)室温屈服强度的方法,以经验参数作为输入变量构建了一个多元线性回归模型,该模型在拟合优度(R2)方面表现出色,达到了0.996,同时模型预测的室温屈服强度的平均误差仅为4.2%。此外,通过皮尔逊相关系数检验,筛选出对室温屈服强度影响最显著的经验参数,包括价电子浓度、理论熔点、混合焓以及熵焓比。

A method to predict the room temperature yield strength of HEA was proposed based on empirical parameters.Taking empirical parameters as input variables,a multivariate linear regression model was constructed,which performed well in terms of goodness-of-fit(R2)at 0.996.Meanwhile,the average error of predicted yield strength is only 4.2%.Moreover,the empirical parameters exhibiting the most significant influence on yield strength were identified through Pearson’s correlation coefficient test,including valence electron concentration,theoretical melting point,mixing enthalpy and entropy-enthalpy ratio.