Design of a Multi-Detection System of Process Abnormal Signals and Its Application in the Chemical Process Industry

Abstract

Technologies of detecting process abnormal signals in fault diagnosis using multi-detection systems for industrial use have been developed in an industrial petrochemical plant. In this paper, the design and application of a multi-detection system using an adaptive digital filter, wavelet analysis, attractor analysis, Bayesian statistical inference, white noise testing and hypothesis testing are presented.
The future tasks of the multi-detection system design of process abnormal signals for industrial use are discussed.