Development of a non-linear regression model for iron ore sintering

Abstract

A mathematical model is developed to study the functional relationship between the strength of sinter and some important parameters such as basicity, coke breeze, moisture and return fines. The correlation developed, employing Brandon's method of multiple regression technique, has been validated through experimental data. The adequacy of fitness of data to the model developed has been tested by Chi-square Test and Correlation Coefficient. The computer package gives the composition of raw mix suitable for desired condition and strength of finished sinter. It also makes possible to determine the levels of operating variables to have the desired quality of sinter.