A time-encoded signal-processing approach to fault classification of an electrohydraulic pressure control system and its application to a hot steel strip rolling mill

Abstract

A technique is presented whereby time domain signals may be classified in terms of four shape parameters. These parameters are processed to create a condition indicator, its variation being used to identify system faults. In this study, the leakage flowrate and a faulty servo-valve are identified on a pressure control system and a comparison is made with a conventional fast Fourier transform approach and a commercial software package, time-encoded signal processing and recognition. A new code, CARCODE, is then applied to 28 servo-valve-controlled work roll bending circuits on a seven-stand hot steel strip rolling mill and a simple graphics presentation is developed to aid real-time condition monitoring.