A blast furnace upper suspension material diagnosis method and device fusing timing characteristics and attention mechanism

By integrating temporal features and attention mechanisms, an intelligent blast furnace hanging charge diagnosis model was constructed, which solved the real-time and adaptability problems of blast furnace hanging charge diagnosis technology, realized accurate identification and real-time early warning of hanging charge, and improved blast furnace smelting efficiency.

CN122153550APending Publication Date: 2026-06-05NANJING IRON & STEEL CO LTD +1
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
NANJING IRON & STEEL CO LTD
Filing Date
2026-01-20
Publication Date
2026-06-05

AI Technical Summary

Technical Problem

Existing blast furnace charge suspension diagnosis technology suffers from insufficient real-time performance and adaptability, making it difficult to effectively learn abnormal features under complex working conditions. Furthermore, the model's accuracy and generalization ability are insufficient when abnormal samples are scarce.

Method used

By integrating temporal features and attention mechanisms, and collecting multi-source historical monitoring data, preprocessing and feature selection are performed to construct a blast furnace overload diagnosis model with a temporal convolutional network, a bidirectional gated recurrent unit network, and an attention mechanism. The model is then trained and evaluated.

Benefits of technology

It achieves accurate identification of suspended materials on the blast furnace, improves the intelligence and objectivity of the identification process, has good engineering applicability and real-time early warning capabilities, reduces unplanned shutdowns and production fluctuations, and improves smelting efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122153550A_ABST
    Figure CN122153550A_ABST
Patent Text Reader

Abstract

The application discloses a blast furnace upper suspension material diagnosis method fusing time sequence characteristics and attention mechanisms, relates to the technical field of blast furnace burden movement state monitoring, and comprises the following steps: collecting multi-source historical monitoring data and performing pretreatment; calculating derived data based on the multi-source historical monitoring data; extracting input characteristic variables representing upper suspension materials based on Spearman correlation coefficients; constructing an upper suspension material diagnosis data set containing normal samples and abnormal samples; establishing an upper suspension material diagnosis model fusing time sequence characteristics and attention weighting, and training the model; performing real-time diagnosis on the blast furnace operation state based on the upper suspension material diagnosis model, outputting suspension material identification results, and comprehensively evaluating the model performance by using a test set. The application integrates time sequence characteristic analysis and attention mechanisms, constructs a blast furnace upper suspension material special diagnosis model fusing time sequence characteristics and attention mechanisms, and effectively improves the intelligent level and objectivity of the identification process.
Need to check novelty before this filing date? Find Prior Art