Mixed Modelling For The Thermal Parameter Of Pellet Mine Based On Mechanism-Neural Networks

As the blast furnace material, the pellet should meet following three-aspect standard: physical performance, chemical characteristics and metallurgical characters. Physical performance is an important indicator of its quality, and compressive strength has the most representative. Almost all the compressive strength texts have not finished till some times after the production had done. And this brings many problems, such like the long time and long cycle of acquire data, and detrimental to the quality control of produce. Even worse, it can hardly recovered if the quality is found unqualified, and will brings baneful influence on quality and performance of enterprises. So it is important to make the judgment of pellets’ quality and establish the prediction model of pellets’compressive strength.The coal is the only reliable sources of heat which is closely related with the temperature in chain grate-kiln process, in the roasting process of pellets. Because of the processes are complicated, serious couplings among the parameters of the processes and operations exist and many uncertainties are unavoidable, it is not easy to grade the optimal control of temperature by the mathematical model. Therefore, how to find out the relations between the coal injection and the temperature in order to save the superfluous energy has become a challenging topic.This paper analyse the technology process of pellet production deeply firstly, and emphasis the temperature in chain grate-kiln system which affects compressive strength of pellets. Then establish the quality prediction model based on Case-Based Reasoning with the temperature of preheat one and two, the temperature of head and end of kiln as the inputs, while the compressive strength as the output. After emphasis the temperature in roasing process which affects the coal injection, establish the temperatures’optimal control model based on Case-Based Reasoning. The inputs are temperature of head and end of kiln and the 2th temperature of east and west copper while the output is coal injection.This paper based on the project from Ansteel, developping quality prediction and temperature optimized control software interface by C#.NET. According to the standard of OPC, it can solve the problem of data transmission from WinCC to C#.NET. And using ADO.NET to link C#.NET to the database for save, search and update the data. The system is proven effective and valuable as reference for guide to the production process after have been used for some times in the factory.