Rotating Machinery Fault Characterization by Recurrence Plot: A review

Abstract

The dynamic behavior of a rotating machine is a reflection of its design, operating conditions, type of application, and operational condition of its components. Through vibrations analysis, the machinery condition is diagnosed, and failures are predicted in advance; however, traditional methods for diagnosing machinery failure treat all dynamic phenomena as stationary and linear regimes, but worldwide standardization and technological advances demand better diagnosing methods to reduce maintenance costs and increase the useful life of machinery. This paper shows that Recurrence Plot is a better method for measurement and fault diagnosis in rotary machinery because nonlinearities and non-stationary regimes are included in the vibration analysis, thus diagnosis and quantification faults can be more precise.