Optimal control of the argon-oxygen decarburising steelmaking process

The argon-oxygen decarburising process is an increasingly important method of stainless-steel manufacture. In the paper, a nonlinear dynamic model is presented which represents approximately the refining process. Finite-dimensional-optimisation techniques are used to calculate an optimal switched control, and infinite-dimensional-gradient methods are applied to discover an optimal continuous strategy. Control is achieved by varying the ratio of argon and oxygen in the supplied gas. Results with a realistic cost function indicate that substantial improvements in operating practice may be achieved. The feasibility of implementing the suggested control schemes is discussed.