Fault diagnosis for hydraulic gauge system based on data fusion

Abstract

Data fusion method is applied in the diagnosing the faults of the hydraulic screw down system of rolling mill, where the difficulties lie in the difficulties of lacking enough theoretical basis and getting the degree of fault measure. In full consideration of the working status and environment of rolling mill, the dynamic time-domain analysis and ARX model were used in analyzing the fault signal data of the rolling mill. By introducing a reasonable measure as the foundation for the application of data fusion theory, using the above two fault diagnosis methods and the subjective experience, faults diagnosis are carried out by data fusion method. The results overcome the limitation and the instability caused by single diagnosis method. The diagnostic results are more reliable. Finally, the effectiveness and accuracy of the method are validated by examples.