The Design On Temperature Prediction Model Of Molten Steel During LF Refining Process

That everyone is familiar with LF,but LF is a widely used electric heating function of the completion of one of the ways,because the casting and the steel-making process is important between the heating and tempering steel.Heat adjusting of the steel water is important work prefaces of Continue Case. The LF electricity heating completes to adjust the steel water temperature, which is used more. In heat process, steel water temperature's control is key to produces the qualified steel, for this reason temperature measure seem to be importance, especially, continue to measure the steel water temperature. Because the steel water temperature degree is general all at 1520~C to 1650℃.Currently, it is very difficult that measure under the so high temperature when work is continuing, even it can not do. For this, people are working hard to look for the material of measure temperature continuously, at the same time, and also at study the different measure method. Among them, according to LF refinement process, the system with steel water, metal alloy, it's the energy income analyzes with the losing system, in according to the heat equilibrium regulation, establish the steel water heats predict module. Because the LF inside steel water temperature is non-line changing and much affect the factor, make this kind of method contain very big localization: The calculation sophisticates and the weak of predict ability. In the last few years, the artificial intelligence technique flies to develop, every kind of applied system piles up one after another. Make use of the artificial intelligence technique in metallurgy profession, also got the metallurgy the technical personnel interest. The artificial intelligence can imitate the person's brain behavior, having the good differentiating and analyze to the complicated behavior.With the model being used to off-line and online forecast for data gathered from spot, results shows that precision of forecast is relatively high average error is less than 5°C. Errors mainly distribute in area from 3°C to 5°C.This context set mold of the T-S fuzzy nerve network, in according to hot equilibrium mathematics of the temperature inside the LF. In operation circumstance that analyze the some steel factory the spot with the foundation of the craft process, make sure The network construction combine estimate model that established the steel liquid temperature. Model network contain 7 importations with an exportation.Proceed the verification to the network model while supposing the constant circumstance in some factors,the network model can reflect the true circumstance of the craft. Suppose, at other factor constant, the power enlarges with electric current electric voltage atemperature for predicting goes up, cooling off the water current with blow the amount of argon become big, a temperature for predicting lower etc. In addition, the model makes use of the language of vb realizes installing to control in the work on board puts to LF whole processes into practice temperature forecast.