Research On Sintering Optimization Iron Ores Matching Integration Technology System

At present,lower cost and higher benefit has become fashion of iron and steel industry. The kinds and proportions of iron ores for sintering are adjusted constantly with change of global iron ore resources. It is significant to find a ore matching scheme, which sintering yield and quality is high while cost is low.Two methods to obtain a proper iron ores matching scheme were put forward, as well as a system framework consisted of iron ore capabilities database, iron ore matching calculation model, BP neural network model, predictive model of sinter quality and optimizing iron ores matching model by analyzing characteristics of iron ores matching in sintering process. Database of iron ores capabilities which consisted of basic capabilities of raw material and sintering capabilities was established by using database technology. BP neural network model was built, and BP algorithm was studied. An improved BP algorithm-Vogl quick algorithm was put forward as BP algorithm has some problems such as low study efficiency, slow convergence rate. Predictive model of sinter quality was built using neural network technology. This model has a predictive rate of over 83.3%, and with good generalization and self-adaptation abilities. Optimizing iron ores matching model was built using genetic algorithms technology.This model can give the best iron ores matching scheme ,which cost is lower and quality of sinter is good . This model has a good optimizing effect. Sintering iron ores matching calculation model was built with which iron ores mixing calculation and proportioning of sintering material calculation were realized. This model can give a proper iron ores matching scheme and material proportioning scheme. The optimizing matching ores system model has been developed using a mixing programming mode of Visual C++ 6.0 development tool and Matlab programming language. This software has some advantages such as good practicability, friendly interface, convenient operation, powerful functions of storage, calculation, prediction and optimization.