Hydraulic cartridge valve group reliability evaluation and fault diagnosis based on data and model integration

The hydraulic cartridge valve group (HCVG) stands as an intricately integrated, high-capacity flow device extensively employed in hydraulic transmission systems. Nonetheless, challenges concerning the dependability and fault diagnosis of the HCVG pose significant ramifications for the overall performance and safety of hydraulic systems. In response to these challenges, this endeavor devises a comprehensive method package to evaluate the reliability and diagnose faults within the HCVG. This method package encompasses two phases: the design and usage of the HCVG. During the design phase, our methodology assesses the HCVG reliability through conditional probability, meticulously accounting for each component’s failure modes and associated effects. Correspondingly, during the usage phase, the methodology encompasses three pivotal steps: state observation, fault diagnosis, and fault location. (1) In the state observation step, we employ an integrated data and model approach that fuses measured data with mathematical models to ascertain the true system state. (2) In the fault diagnosis step, we utilize an adaptive threshold technique, dynamically establishing thresholds based on discrepancies between observed and actual states. (3) In the fault location step, we harness the Graph Convolution sensitivity method, quantifying each component’s impact on reliability through Graph Convolution and pinpointing the faulty component exhibiting the highest sensitivity. Finally, the above method package undergoes validation via simulation and experimentation using a typical HCVG scenario, conclusively demonstrating its capacity to effectively and accurately diagnose and pinpoint faults within the HCVG. This undertaking holds substantial significance for enhancing the reliability and fault diagnosis of the HCVG and other hydraulic systems.