| Abstrakt: |
Predicting the remaining useful life of hydraulic pumps is critical for ensuring their optimal performance and extending their operational lifespan. Accurate remaining useful life predictions enable timely adjustments to the pump's working conditions, thus enhancing maintenance strategies and preventing unexpected failures. This review comprehensively examines the two primary categories of remaining useful life prediction methods for hydraulic pumps: data-driven and model-driven methods. Data-driven approaches leverage historical and real-time operational data, employing machine learning and statistical analysis techniques to forecast remaining useful life. In contrast, model-driven methods utilize physical models and failure mechanisms to predict the remaining lifespan based on the pump's working conditions and inherent characteristics. By evaluating the strengths and limitations of these methods, this review aims to offer insights into their practical applications and future research directions in the field of hydraulic pump prognostics. [ABSTRACT FROM AUTHOR] |
