Study On The Modle Of Breakout Prediction In Continuous Casting Process

Breakout is the most detrimental event in continuous casting. To develop tenchniques of the breakout prediction is very effetive to prevent breakout, it could be valuable in theoretical research as well as the societal economy. In this paper, reference to the results of researchers at home and abroad, study the breakout in slab continuous casting. Study content are as follows:①Discuss the sticker type breakout which is the most common type of breakouts from mechanism of the breakout in detail and made a prediction measures.②Introduced the principle of thermocouple temperature measurement in breakout prediction, analysis the data which came from the steel plan.Then analysis the prediction model based on logic judgement.③Established a wave of temperature ratio to improve the former breakout prediction models which can not monitor the hot spot propagation well.④In analyzing the prediction model based on logic judgement, with the mechanism of breakout, a prediction model based on BP neural network was established. The model was built by shift register, time neural network which used to identify the mode of temperature change and wave of temperature ratio ,the buffer register and space neural networks which used to determine whether the breakout would happen.⑤Analysis the data collected in steel plan, chose the breakout and normal samples to train the prediction model.Based on the reseach of above, develop the programme of time nerual network, space nerual network, training and testing programme. Test results show that the prediction based on BP neural network can predict more effectively and quickly than the the original system in the plan which base on logic judgement.