Application of Neural Net Modeling and Inverse Control to the Desulphurization of Hot Metal Process

Optimization techniques are powerful in determining inputs to a process that will drive its output to a desired target. This inverse control problem reduces to minimizing a positive cost function that measures the difference between the output and its target value. In this paper, we present a method for inverse control that uses a combination of artificial neural net modeling and optimization. We apply this method to the desulphurization of hot metal process. In this steel industry application, the sulphur content of hot metal, obtained at the end of calcium carbide powder injection into 400 ton torpedo ladles is predicted as a function of hot metal weight, treatment time, initial sulphur content, gas flow rate, and powder injection rate. Based on the prediction model, the optimization algorithm coordinates the five inputs or part of them to achieve a desired sulphur content in the hot metal.