Statistical experiments using the multiple regression research for prediction of proper hardness in areas of phosphorus cast_iron brake shoes manufacturing

International Conference on Applied Sciences 2017

Kiss, I.^1 ; Cioat, V.G.^1 ; Ratiu, S.A.^1 ; Rackov, M.^2 ; Peni, M.^2

University Politehnica Timisoara, Faculty of Engineering Hunedoara, Department of Engineering and Management, Hunedoara, Romania^1

University of Novi Sad, Faculty of Technical Sciences, Novi Sad, Serbia^2

关键词: Chemical compositions;    Independent variables;    Mathematical solutions;    Multiple correlation;    Multiple linear regression models;    Multiple regressions;    Regression coefficient;    Statistical experiments;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/294/1/012078/pdf
DOI  :  10.1088/1757-899X/294/1/012078

来源: IOP

【 摘 要 】

Multivariate research is important in areas of cast-iron brake shoes manufacturing, because many variables interact with each other simultaneously. This article focuses on expressing the multiple linear regression model related to the hardness assurance by the chemical composition of the phosphorous cast irons destined to the brake shoes, having in view that the regression coefficients will illustrate the unrelated contributions of each independent variable towards predicting the dependent variable. In order to settle the multiple correlations between the hardness of the cast-iron brake shoes, and their chemical compositions several regression equations has been proposed. Is searched a mathematical solution which can determine the optimum chemical composition for the hardness desirable values. Starting from the above-mentioned affirmations two new statistical experiments are effectuated related to the values of Phosphorus [P], Manganese [Mn] and Silicon [Si]. Therefore, the regression equations, which describe the mathematical dependency between the above-mentioned elements and the hardness, are determined. As result, several correlation charts will be revealed.