The Research Of Sintering Optimal Matching System

The source of iron ore has changed a lot when the domestic iron ore can't meet the demands of Iron&Steel making .The proportion of overseas iron ore has improved incessantly and the number of quarry is increasing. In order to keep the quality of sintering cake, we must make full use of the technology .It's a new challenge in iron ore selecting and ore matching and the more flexible sintering yielding.When the sort and the mass of iron ore have changed a lot, how can we boost the index of the sintering yield and sintering produce quality and reduce the matching cost of iron ore and improve the industrial international competence? To acclimatize ourselves to the frequent changes of iron ore, how can we get multitudinous matching project? All of these are what Iron&Steel industry cares about. In the last analysis, it concentrates these into optimal matching.In this paper, we have analyzed the matching status quo of sintering yielding Iron&Steel industry. Based on the property of matching, a optimal matching model (OMM) is proposed to solve the difficult and challenging problem of sintering matching. With this model, we can solve the difficult and challenging problem of optimal matching. OMM is constructed by iron ore chemical composition restricted model (CCRM) and the optimal model of sintering cake yielding and mass (YMOM) .Matching Simulate Model (MSM) has been involved in YMOM.Based on the property of matching data and the shortage of BP neural network, a integrative neural network (INN) is proposed. The results obtained from the actual prediction demonstrate that the performance and capability of the proposed INN are superior. With the appliance of INN, we construct MSM from the vast given input and output data to predict the index of yield and quality of sintering cake. With the appliance of genetic algorithm (GA), we construct YMOM simultaneously to produce multitudinous matching project. We name it hypo-optimal matching project which make it possible to get multitudinous optimal matching project.Adopted such new technologies as object-orient , visual programming , ADO database interface and mix programming, a sinteringoptimal matching software system has been developed in this paper, which facilitate the transfer of technology emerging from theoretical research of optimal matching into industrial applications.The software system consists of three parts-database managing, integrative neural network modeling platform and optimal matching platform, which are independent and secure. It can be applicated to different sintering plant and maintained easily.