Dynamic Forecasting Model Of BFG And Its Application In Iron And Steel Works

BFG is the by-product of the blast furnace iron-making, which is one of the important gas resources of iron and steel works, and directly impact on the effective use of energy consumption index of the enterprise. It could significantly reduce the gas diffuse and reduce energy consumption of the enterprise for distribution gas and scheduling to provide timely and accurate forecast information to establish proper forecast model of BFG production, consumption and storage. This paper aim at the dynamic forecast problem of BFG production, consumption and storage, specific studies are as follows:(1) Start with steel production technology and BFG production equipment running status, to deeply discuss the influence of BFG when the equipment running status changed and combined with the raw material conditions, thermal efficiency, thermal operation to find out the influence iactors of affect the gas producing and dissipation capacity, for theoretical foundation for the BFG dynamic prediction model.(2) The establishment of the Mix_Bp_lssvm prediction model that a higher prediction precision and the Updata_Lssvm prediction model which can run online. For that the prediction accuracy of the traditional model is not high of the actual situation, this paper uses the wavelet analysis method to preprocess the blast furnace gas amount of historical data, and separate the trend and wave data. Then, combining with the characteristic of Bp and Lssvm model prediction, the paper establishes Mix_Bp_Lssvm prediction model with higher accuracy; At the same time, in view of the traditional forecast model can not solve the problem of variable working condition of gas prediction, the paper uses recursive, weights and thresholds updated to improve Lssvm forecasting model, and builds a Updata_Lssvm prediction model with function of time series updating and self-correcting. Through the actual operation data validation to prove:iron and steel enterprise Updata_Lssvm prediction model can forecast the data in variable working condition, and it has good prediction accuracy, and its computation time is moderation, and it can forecast the gas variation online.(3) Combined with vector space transformation theory, the paper uses the grey correlation analysis method to eliminate the small users, simplify the blast furnace gas system constructing and the gasometer influencing factors, establishes blast furnace gas buffer unit gas consumption and storage capacity prediction model. Through the blast furnace gas dynamic prediction system verification proves that the model has better prediction effect.(4) According to the blast furnace gas dynamic forecasting model, the paper using Matlab develops blast furnace gas dynamic prediction system, and realized the dynamic prediction of blast furnace gas production, consumption and buffering capacity. At the same time, it provides a platform to analyze the practicability and validity of model in this paper. Combined with the actual gas history data in iron and steel enterprises, the paper gives a applicable case of blast furnace gas dynamic prediction system.