Significance Statement
Modern facilities, especially within smart manufacturing environment, are massively instrumented with IP-enabled intelligent sensors and controllers gathering on-time process information at a high sampling rate. These devices produce abundant process data beyond the capacity of conventional analytical tools based on first-principles models and facilitate the development of data-driven methods. One application of data-driven methods is to detect and diagnose plant-wide oscillation problems often faced by process engineers.
This paper addresses the challenge of plant oscillations caused by poorly tuned controllers; actuator nonlinearities or valve stiction at one point often propagate to the whole operating unit or the whole plant because of a highly integrated system, leading to inferior quality products or even plant safety issues. Conventional physical modeling tools cannot extract useful information from abundant plant data on quickly locating the root cause of oscillation and the relevant variables affected along propagation routes, especially with the interference of noise.
In this paper, the authors proposed a novel framework within the frequency domain to provide systematic diagnosis, which consists of steps like spectral envelope-based band selection, spectral transfer entropy based cause-effect analysis and root cause validation by wavelet power spectrum. Test cases based on simulated data and industrial data provided by Eastman Chemical Company are provided.
The results demonstrate successful applications of such a framework: the common frequency reflecting abnormal oscillations and associated contributing indicators, cause-effect relationship between those variables and the resulting process topology within frequency band selected, and the oscillation occurring time can be all visualized intuitively. The negative influence of noise coming from contaminated raw plant measurements is mitigated.
In summary, the frequency-based methodology framework presented in their research provides a convenient and promising data-driven tool to overcome the plant-wide oscillation detection and diagnosis challenge faced by process engineers.
Journal Reference
Root Cause Diagnosis of Plant-Wide Oscillations Based on Information Transfer in the Frequency Domain
Ind. Eng. Chem. Res., 2016, 55 (6), pp 1623–1629.
Shu Xu1, Michael Baldea1, Thomas F. Edgar1, Willy Wojsznis2, Terrence Blevins2, Mark Nixon2
[expand title=”Show Affiliations”]- McKetta Department of Chemical Engineering,The University of Texas at Austin, Austin, Texas 78712, United States
- Process Systems and Solutions,Emerson Process Management, Round Rock, Texas 78759, United States
Abstract
Plant-wide oscillations generated in a single unit can negatively affect the overall control performance of the process; thus, it is necessary to detect them and diagnose their root cause. However, the interference of noise and the need for oscillation propagation routes pose more challenges for process engineers. In this paper, the concept spectral transfer entropy is proposed and its connection to the spectral Granger causality is derived. Moreover, an information transfer method incorporating spectral envelope algorithm and spectral transfer entropy is applied to provide new diagnostic guidance, whose feasibility and effectiveness have been demonstrated by both simulated and industrial case studies. Compared with current methods, the new procedure enjoys the following advantages: (a) performing oscillation detection and diagnosis within a targeted frequency range and mitigating the effects of measurement noise outside the bandwidth; (b) provides an nominal causal map reflecting the oscillation propagation pattern. The root cause obtained by the method in the industrial case is further validated by a wavelet power spectrum.
Copyright © 2016 American Chemical Society.
Go To Ind. Eng. Chem. Res