Study And Application Of The Smelting Furnace's Control Strategy On The BP Neural Network

Copper matte converting is an important process in copper pyro-metallurgy, involving chemical reactions, fluid flow and transformation of both heat and mass. It is characterized by properties such as multi-variables and non-linear, couple changes with great inertia and uncertainties, etc. Therefore, the abnormities of copper refining furnace always happens. Because in the feeding process, the materials differ greatly and the various factors influence the process amazingly, the online control of the converting process is always difficult to realize. The product of the matte converting process is copper. The aim and key of the process are to prevent over as well as under-converting and guarantee the quality of the copper. Therefore, in order to make full use of P-S converters' productive efficiency and try hard to decrease the influences of uncertain factors artificially to further strengthen produces, save energy and reduce consumes, it is more and more important to carry out modeling simulation research on copper matte P-S converters.The aim to optimize decision on copper matte converting process is to improve productivity, achieve high quality yield and reduce energy consumption. In order to meet the demands of enhanced production, coordinate better systematic arrangement and tap the latent power of system production scientifically, the development and survey of the of P-S converters' end-point control technology are analyzed in this paper seriously. Based on comprehensive study of the copper matte converting productive process, a calculative model for material balance of metallurgy has been set up, calculation of theoretical productivity of copper and ratio-utilized in converting process has been carried out, the recently developing artificial neural network technology has been taken use of ,and the static prediction model of end-point content for oxygen-converter has been established and take full advantage of the artificial neural network technology developing recently which has established the static prediction model of end-point content for oxygen-converter based on BP neural network with MATLAB math work as simulation software.In the process of modeling, the author makes use of the function about BP neural network toolbox attached in the MATLAB and the procedure written according to the basic principle of BP neural network separately to carry on assistance judgment on the converter's end-point in the building the copper period. Thus, this way can improve accuracy and practicability of the model greatly, prevent and reduce the quality accidents in copper matte blowing effectively, achieving the goal of saving the energy consumption.According to the analysis of the data of some month in Guixi smelter and refinery of nonferrous metals converter factory, this model applied to its production better, has certain instructional function in terms of the production and certain theoretical and practical values.