Static Control For Converter Smelting Operation Based On Neutral Network

In recent years, artificial neural network technique develop fast and fierce, and get increasingly extensive research with apply. It can be used for operation and the process control of production in the metallurgical industry. This paper introduced some correlative data of nerve networked in brief, and the paper adopted the now technique of popular BP neural network. The 30t converter's hit rate of endpoint temperature and carbon content of the iron and steel company of Xingtai is low. Aim at to the problem, we generalize 29 parameters that affect the endpoint temperature and carbon content and developed a set of expert system with Visual Basic6.0. The core of the expert system is a static model based on artificial neural network.This paper choose the actual data of continuous 300 batch from XG as example, after screen sizing,normalization , then established book data base. According to the endpoint temperature and carbon content, we improve the BP algorithm and establish two three-layers BP prediction neural networks. The two networks individual predict the endpoint temperature and carbon converter .And we get a good result when we put the model into practice in converter mill of XG.This paper considered that in the metallurgical factory of china, especially Hebei province, the most converters are medium and small-sized converter. The converter can't insert the Para- spear, so it can't use the dynamic control technique. The normal static model's calculation accuracy is bad, target hit rate is low, and it's application effect in practice is not good. The static artificial neural network just fill up the blank, .it isn't necessary to modify the existing equipment for implement the computer open loop control. This technology will play a very important role in cost reduction,standard operation and stabilize product quality.