Detection and Diagnosis of Oscillations in Process Plants

摘要

An Wavelet analysis tool is utilized for the detection of oscillations in actual industrial chemical plants. It can detect multiple oscillations in measurement signals and shows valuable information for the diagnosis. For each of the detected oscillatory control loops, the root cause of the oscillation is diagnosed by the analysis of input-output characteristics of a valve. Finally, the combination of the Wavelet tool and the analysis of MV-PV plots is shown to be a powerful practical method for the detection and diagnosis of oscillations in process plants.