Research On Oxygen Consumption Prediction And Scheduling Optimization In Iron And Steel Industry

At present,the unbalanced supply and demand of oxygen has been plaguing iron and steel production enterprises.It will not only cause unnecessary waste of resources,but also affect the normal production and operation of enterprises.This paper is based on the software and hardware platform of the intelligent energy management and control platform project of a Steel company,aiming at oxygen consumption prediction and balance optimization scheduling model.This paper focuses on the analysis of the influence factors of oxygen consumption on the main oxygen consumption equipment,and establishes the oxygen consumption prediction model using neural network algorithm.At the same time,according to the pressure of oxygen pipeline and the volume of Oxygen Spherical tank,the oxygen production load adjustment scheme is formulated,and the oxygen consumption prediction model is finally formed.The cooperation between production plan and energy plan is achieved in a real sense,and solve the oxygen supply and demand balance model constraints complex problems.The main contents of this paper are as follows:(1)Firstly,the important factors affecting oxygen consumption are analyzed,and the data acquisition range is determined.Data communication interface is established with related systems at all levels to integrate production data and energy data.Then,the important influencing factors related to oxygen consumption in blast furnace ironmaking and converter steelmaking are further determined by correlation analysis technology,and the key input variables are determined.The data are cleaned and sorted to form sample data sets.(2)The BP neural network algorithm and time series algorithm are selected as the theoretical basis of the system,and the oxygen consumption prediction model is developed by using IBM SPSS data mining software.The oxygen consumption of blast furnace and converter is predicted.On the basis of full analysis of oxygen consumption of oxygen-consuming equipment converters and blast furnaces,oxygen storage equipment network and oxygen buffer capacity of liquid oxygen,combined with the characteristics of steelmaking process planning and oxygen consumption of blast furnaces,the oxygen prediction model of converter based on BP neural network algorithm and oxygen consumption prediction model of blast furnace based on time series algorithm are established.By evaluating and comparing the accuracy of the models,the prediction accuracy of the models built by the two algorithms can reach more than 99% in 4 hours,which meets the needs of practical oxygen pipeline network equilibrium optimization application.(3)According to the different objectives of optimization,using the oxygen consumption predicted by the oxygen prediction model as input,different oxygen pipeline network equilibrium optimization models(including the minimum oxygen emission and the minimum unit consumption of oxygen generating units)are established.The results of different models are analyzed and compared,and the dynamic selection and dynamic optimization of optimization models are carried out according to the actual business requirements.This system comprehensively considers various influencing factors of oxygen consumption in blast furnace ironmaking and converter steelmaking,which covering a comprehensive range,therefore,the conclusion is more instructive and practical.Through the establishment of the model,this paper initially solves the problem of imbalance between supply and demand of oxygen system in steel enterprises and the inaccuracy of results based on manual experience scheduling methods,which plays a key role in reducing energy costs and improving energy efficiency.