Scholarlink

<p>Breakout is an extremely serious accident in continuous casting, which not only damages the billet and reduces the service life of the equipment but may even cause casualties. To capture and identify the temperature timing characteristics of thermocouples during steel breakout, K means clustering centers were optimized using a genetic algorithm based on the large difference in temperature characteristics between normal condition and breakout condition. The temperature samples of different conditions were aggregated and separated, and then input to the Transformer-BiLSTM model for feature identification and classification prediction. Finally, network model based on deep learning clustering was developed and applied to the field of continuous casting breakout prediction systems. The breakout prediction model was tested in conjunction with actual continuous casting production data. The results show that the accuracy of this continuous casting breakout prediction system is 99.5% and the reporting rate is 100%.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Steel Breakout Prediction System Based on Deep Learning and Clustering

  • 期刊论文

  • 发表日期：

    2025年1月3日

查看全文
  • Benguo Zhang,
  • Heng Wu,
  • Haochen Yu,
  • Ruizhong Zhang,
  • Lifeng Fan

摘要

Breakout is an extremely serious accident in continuous casting, which not only damages the billet and reduces the service life of the equipment but may even cause casualties. To capture and identify the temperature timing characteristics of thermocouples during steel breakout, K means clustering centers were optimized using a genetic algorithm based on the large difference in temperature characteristics between normal condition and breakout condition. The temperature samples of different conditions were aggregated and separated, and then input to the Transformer-BiLSTM model for feature identification and classification prediction. Finally, network model based on deep learning clustering was developed and applied to the field of continuous casting breakout prediction systems. The breakout prediction model was tested in conjunction with actual continuous casting production data. The results show that the accuracy of this continuous casting breakout prediction system is 99.5% and the reporting rate is 100%.