Fault Diagnosis and Analysis of Hydraulic System Based on Partial Least Squares Method

The hydraulic system is widely used as power mechanism and actuation system in large-scale industrial equipment. The strong coupling between system circuits and different fault mechanisms make fault diagnosis difficult. In order to address these issues, a fault diagnosis method based on partial least squares is proposed for hydraulic system. The AMESim simulation platform is used to simulate the fault status of hydraulic components, generate fault data, and combine partial least squares to establish a regression equation between process variables and fault types for discriminative analysis of hydraulic system fault types. The simulation results indicate that the partial least squares method can achieve process variable data fusion, extract fault features, and complete the detection, separation, and recognition of fault types.