| Abstrakt: |
The objective of this study is to identify the influence of the chemical composition of alloyed steel on the tensile strength through predictive multiple regression models of the first and second degrees. Data on the percentage content of nine chemical compounds in alloy steel are used as independent variables: C, Cr, Mn, Mo, Ni, P, S, Si, Al, and the product's diameter. In order to accurately perform multivariate linear regression, all variables undergo Yeo-Johnson transformation in advance to achieve normal or near-normal distribution. The obtained linear regression models fit the measured values for tensile strength with 76% and the models with predictors up to second order – with 97.6%. The alloy compounds with the strongest influence on tensile strength are identified. [ABSTRACT FROM AUTHOR] |
