Chatter Monitoring and Diagnosis for High-speed Cold Rolling Mills

Abstract: Chatter is particularly prevalent in the high-speed cold rolling mill production process, which has a negative impact on product quality and efficiency. This research combines industrial data, including chatter data and production process data, and applies the ICA (Independent Component Analysis) approach to build a chatter monitoring and diagnosis model for a 1420 high-speed cold rolling mill. The proposed model can realize the prediction and diagnosis of cold rolling chatter. When the abnormal chatter is monitored, the model uses the contribution rate of each process variable to diagnose the reason of the abnormal chatter. The model was applied to analyze actual production data. The results of research and analysis show that the proposed chatter monitoring and diagnosis method based on ICA can accurately predict the abnormal chatter of the rolling mill, and can analyze the cause of the abnormal chatter.