Abstract

PROAD is an automated framework developed using multi-pronged approach that integrates process, vibration, and spectral data for performance monitoring, prescriptive maintenance, and fault diagnosis of centrifugal pumps. Data input to PROAD is through robust interactive templates. A clustering algorithm is implemented within PROAD that computes deviation of pump head (as a function of flowrate) w.r.t. vendor-supplied head curve for fresh pump. Head deviations over time are analyzed using regression techniques to predict remaining useful life of pump w.r.t. threshold limits. Data from vibration sensors are analyzed to prescribe spectral checks and data obtained from spectral checks are then examined to diagnose mechanical faults. Based on performance and diagnostic metrics, color-coded health status, report, and plots are displayed for each pump. We have tested PROAD on industrial data and it successfully diagnosed a pump with severe bearing damage. PROAD can be easily implemented in industries for health monitoring of centrifugal pumps.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call