Open AccessA Prediction Model of Alloy Yield in RH Furnace Based On SSA-ELM
Author(s) -
Jiaqing Guo,
Huikang Liu,
Lin Chai
Publication year - 2022
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2216/1/012097
Subject(s) - yield (engineering) , artificial neural network , alloy , convergence (economics) , stability (learning theory) , process (computing) , range (aeronautics) , mean squared prediction error , algorithm , biological system , computer science , materials science , artificial intelligence , machine learning , metallurgy , composite material , economics , biology , economic growth , operating system
Aiming at the problem that it is difficult to predict the alloy yield in the RH furnace refining process, an alloy yield prediction model based on a sparrow search algorithm (SSA) optimized ELM neural network is proposed. Firstly, because the dimensions of input parameters are different, 11 input features are reduced by factor analysis (FA), and 5 input features are obtained. Then, through the sparrow search algorithm with fast convergence, high precision, and good stability, the input weight value and threshold of the ELM neural network are optimized, the SSA-ELM alloy yield prediction model is established, the alloy yield is predicted, and the off-line operation of the model is realized, which provides a theoretical basis for the prediction of alloy yield in the actual production process. Finally, by comparing the simulation results with the actual production data, it is found that the prediction results of the target element alloy yield predicted by the SSA-ELM are within the error range, and the prediction accuracy is higher than that of the ELM prediction model, which verifies the feasibility and effectiveness of the model.