The formulas of calculating the strip temperature inside the pickling tank and the heat quantity are established.

The generation and alloction of deformation-heat and friction-heat are marjor factors affecting the strip temperature.

Model of prediction for the terminal temperature of molten steel in LF refining;

Model of Predication for Temperature of Molten Steel in a 60 t Tundish;

Model of Prediction for End Temperature of Molten Steel in CAS Refining;

The model is the basis of controlling the molten steel temperature.

Combining the conventional mechanism model with the newly developed ELM(extreme learning machine),a hybrid model was developed for soft sensing of molten steel temperature with the aim to rise above the complexities of the physico-chemical reaction process and heat transfer process during LF(ladle furnace) refining.

By optimizing the technological regulation,importing real-time monitoring,strengthening the management,the molten steel temperature stability can be assured and the breakout rat.

A neural network prediction system of molten steel temperature for VD has been developed.

The prediction model of molten steel temperature for LF VD CC route has been developed by means of meural network method The difference between the predicted and measured temperature was in range of 0 7 ℃ for 85 % heats,and less than 10 ℃ for all heats Because of data replacement function in the model,the prediction precision can be guaranteed regardless of the influence of gradual change factor