Development of an Operation Guidance System to Reduce Pushing Load of Coke Ovens Based on Probabilistic Optimization

Abstract

In recent years, the increase in the pushing load of coke ovens caused by aging has become a problem in steel works. In order to reduce this problem, an operation guidance system by a statistical approach was developed. The prediction model was developed based on an individual oven database, to deal with the variations in the characteristics of the ovens. The pushing process which involves complicated motion of coke grains can be regarded as a probabilistic phenomenon. Therefore, a guidance system using probabilistic optimization was constructed. The system was evaluated in actual operation, and the effect of reducing high pushing load ratio to 1/3 has been confirmed.