Energy features fusion based hydraulic cylinder seal wear and internal leakage fault diagnosis method

Internal leakage is one of the most common faults in hydraulic cylinders, and seal wear is the main factor in internal leakage. However, it is difficult to detect seal wear and internal leakage in hydraulic cylinder using present approaches due to the complex hydraulic system. Therefore, an intelligent fault diagnosis method based energy features fusion is proposed to detect seal wear and internal leakage. First, computational fluid dynamics (CFD) technology was adopted to analyze the flow field in the internal leakage area of hydraulic cylinder, and it was found that energy features of pressure signal are related to internal leakage. Then, wavelet packet transform is applied to extract energy features of pressure signal. Finally, energy features is decomposed into statistics by multivariate statistics theory. Statistics are used to detect piston seal wear and internal leakage. The proposed method creatively studies seal wear and internal leakage from the perspective of flow field analysis, which does not require a large number of fault samples and complicated parameters optimization. Experimental in-vestigations are performed to validate the performance of the proposed approach. It is shown that the proposed method has much more robustness and accuracy than several classical fault diagnosis methods. The study does provide an effective way to detect seal wear and internal leakage in hydraulic cylinder.

Automated summary

This study proposes an intelligent fault diagnosis method based on energy features fusion to detect seal wear and internal leakage in hydraulic cylinders. By analyzing the flow field using computational fluid dynamics (CFD) technology, the energy features of the pressure signal are found to be related to internal leakage. Wavelet packet transform is applied to extract the energy features, which are then decomposed into statistics using multivariate statistics theory. Experimental investigations confirm the method's robustness and accuracy, outperforming several classical fault diagnosis methods.