A Fault Analysis and Pattern Recognition Method for Typical Components of Complex Systems

This paper aims to solve the problem of lack of high-quality fault data for fault diagnosis of complex hydraulic systems and typical component failure models. Based on root cause analysis, the paper combines the failure mechanism and key parameters of engine drive pumps, electro-hydraulic changeover valves and hydraulic cylinders, which are common components in hydraulic systems. The paper builds simulation models of these components under different test bench conditions, which can generate operational health and failure data under different operating conditions. The paper analyses the flow and pressure data from multiple collection points in the system and components with multiple failure modes and demonstrates that the proposed method can support the fault diagnosis data requirements and provide data support for fault diagnosis and condition prediction of the system and components.