A Novel Fault Diagnosis Method for Multimode Hot Rolling Process Based on MGU and BLS

Owing to changes in raw materials, set points, external environment, and other factors. Hot rolling process (HRP) typically has multiple operating conditions. When the operating conditions change, the performance of the fault diagnosis model may deteriorate. Therefore, effective multimode fault diagnosis technology is an effective way to guarantee the safe operation of industrial systems, and it is also a research hotspot and difficulty in fault diagnosis technology. In this paper, we propose a novel fault diagnosis method for multimode HRP by integrating minimal gated unit (MGU) and broad Learning system (BLS). First, MGU is used to extract the dynamic features of process data. Then, the dynamic features are introduced to the BLS model for fault classification. A case study on the multimode HRP illustrates that the proposed method achieves superior fault diagnosis performance compared to traditional MGU and BLS.