Self-Tuning Multi-Step Prediction Applied to Speed Control of a Sinter Strand

The processing of iron-ore to form sinter is a complex metallurgical process carried out on a moving grate. The production rate of good quality sinter is affected by significant variations in the properties of the raw input materials and in operating conditions. In large production units the emphasis is on overall steadiness of the process and the sinter strand is driven at constant speed for long periods. In principle, the efficiency of the process can be improved by manipulating strand speed, but the control problem is complicated by the high order of time delay involved in the system. The paper presents a case for the application of multi-step prediction methods which can provide the operator with advance information on waste gas temperature variation and guide his action in adjusting strand speed.