Kun Ming university of science and technology
National Natural Science Foundation of China(Research of flame blowing information features extraction andmoltenpool carbon temperature continuous re-al-time prediction model for converter steelmaking endpoint control) （No.61863018）
Continuous real-time prediction of carbon and temperature is the key to the end point control in the process of converter steelmaking. Aiming at the problem that the fluctuation of process data affects the similarity measurement of furnace samples, which causes difficulties in modeling and poor generality, the time series characteristics of steelmaking process data are also considered, a real-time learning method of automatic clustering and calculating the posterior probability of samples to be tested is proposed. Firstly, the fuzzy C clustering strategy weighted by grey relational degree is adopted to automatically cluster the historical database samples. Secondly, the mixed gaussian model is used to calculate the posterior probability of samples to be tested to determine the sample set with the largest correlation degree. Finally, measure out the best small sample a subset of the sample under test, in turn, to predict end point carbon temperature with LSTM network, through the method of data verification, steel converter steelmaking experimental results show that, in accordance with requirements of the steelmaking process, the temperature prediction error on the accuracy of ±10 ℃ is 93.3%, the accuracySof carbon content of the prediction error in ±0.02% is 90.0%.