A Sintering Process Optimization System Based On Neural Network

With the rapid development of the Iron and Steel Industry,the reduce of price causing by the serious over-production capacity in China leads to profits of iron and steel enterprises decline and an increasingly intense competition within the same industry.Therefore,in order to reduce production costs,improve product qual ity,and strengthen their own competitive power,iron and steel enterprises increasingly expect to improve the production efficiency and benefit through controlling method of advanced process.Sintering productions are the raw materials of iron-making production.It directly restricts and influents the output,quality and energy consumption of iron-making production that whether the sintering production is sufficient and whether the chemical composition of material properties can meet the production demand.The sintering ending-heat point has a direct effect on demands mentioned above,and in the actual product the sintering burning-through-point(BTP for short)is mostly used for judging.This paper attempts to set up a predicting model of sintering BTP basing artificial neural network to solve the prediction problem of sintering BTP.This paper manages to set up a predicting model of the sintering BTP basing on the BP neural network to solve the problem of sintering BTP.Through searching and analyzing a large number of domestic and foreign research data and literature,premising on the basis of the current research situation worldwide on position prediction of the sintering BTP,analyzing the basic process of sintering production process and the basic method of setting model on artificial neural network theory.The final four parameters have been determined,which are the flue negative pressure value,the air speed of roller,the advance speed of roller and ignition starting temperature,with deeply analyzin g the production process,basing the requirement that the actual production process can be detected online and the size of the program influencing the sintering final point.On the basis of studying the characteristics of sintering process and BP neural network algorithm,this paper establishes a prediction model based on BP neural network prediction of sintering BTP,and then verifies the network by simulation method.The simulation process firstly is to train the network,and the training samples uses the actual data of produce locale.After the network output achieving acceptable error,the neural network is tested by different data,and gets a conclusion through prediction analysis.BP neural network can be trained within a relatively short period,and the error between the prediction value and the expected value is small.The simulation results show that the method has good effect on solving the problem of the prediction of BTP,therefore,it can be used for the guidance of sintering production process.