A methodology based on cyclostationary analysis for fault detection of hydraulic axial piston pumps

Condition monitoring has been an active area of research in many industrial fields during the last decades, particularly in fluid power systems. This paper presents a solution for the fault diagnosis of a variable displacement axial-piston pump, which is a critical component in many hydraulic systems. The proposed methodology follows a data-driven approach including data acquisition and feature extraction and is based on the analysis of acceleration signals through the theory of cyclostationarity. An experimental campaign was carried out on a laboratory test bench with the pump in the flawless state and in faulty states. Different operating conditions were considered and each test was repeated several times in order to acquire a suitable population to verify data repeatability. Results showed the capability of the proposed approach of detecting a typical fault related to worn slippers. Future works will include tests in order to apply the approach to a wider set of faults and the development of a classifier for accurate fault identification.