Research Of Power Supply And Argon Blowing Model Of LF Furnace Based On Data-driven

Ladle furnace is an important equipment for secondary refining of molten steel in the steel smelting industry.Power supply and argon blowing are the key processes in the refining process.Reasonable power supply system can effectively improve production efficiency,reduce energy consumption loss.Reasonable argon blowing system can promote the uniform fusion of liquid steel components,ensure that liquid steel components meet the standards and meet the quality requirements of steel products.In view of the inaccuracy and instability of power supply and argon blowing system developed by the current iron and steel smelting industry based on mechanism analysis combined with manual operation experience,this paper makes an in-depth study from the data-driven point of view.First,this paper studied the traditional power supply and argon blowing models,determines the influencing factors and sample data of power supply model and argon blowing model by combining correlation analysis method,designs support vector regression(SVR)power supply and argon blowing prediction model based on combination kernel,and establishes comparative models.The evaluation indexes are used to analyze the predictive effect of each prediction model.The result indicates that the predictive effect of SVR prediction model based on combination kernel is better than other models and has practical application significance in small sample data volume scenarios.Then,in order to build a power supply and argon blowing prediction model suitable for large sample data volume scenarios,this paper is base on LSTM,aiming at the different importance of different influencing factors in power supply and argon blowing process at different time series,designs LSTM-Attention prediction model which combines attention mechanism,the evaluation indexes are used to analyze the predictive effect of each prediction model.The result indicates that the LSTM-Attention model has better predictive effect than the LSTM model.In order to verify the validity of LSTM-Attention prediction model in practical application,this paper compares the LSTM-Attention prediction model with the traditional power supply and argon blowing model of a steel plant in China by accuracy evaluation index.The experimental results show that the accuracy of LSTM-Attention prediction model is higher than the traditional model,which indicates that the model has practical application significance.Finally,in order to test the application value of LSTM-Attention prediction model,this paper developed the power supply model and argon blowing model.By collecting production data in the smelting process,tested the functions of power supply model and argon blowing model,and the power supply system and argon blowing system are accurately formulated in production,thus guiding the employees in production.