Abstract
Oscillations are common in closed-loop controlled processes which, once generated, can propagate along process flows and feedback paths of the whole plant. It is important to detect and diagnose such oscillations to maintain high control performance. This paper presents a new data-driven time series method for diagnosing the sources and propagation paths of plant-wide oscillations. The proposed method first uses a latent variable method to select features which carry significant common oscillations, then applies both time-domain Granger causality and spectral Granger causality to provide reliable diagnosis of oscillation sources and propagations. Simulation tests and an industrial case study are shown to demonstrate the effectiveness of the proposed method.
| Original language | English |
|---|---|
| Pages (from-to) | 450-459 |
| Number of pages | 10 |
| Journal | Journal of Process Control |
| Volume | 24 |
| Issue number | 2 |
| Early online date | 3 Dec 2013 |
| DOIs | |
| Publication status | Published - Feb 2014 |
| Externally published | Yes |
Bibliographical note
The authors appreciate the inspiring discussions with Prof. Yan Liu of Computer Science Department at the University of Southern California.Funding
The authors are grateful for the financial support from the China Scholarship Council and the Texas-Wisconsin-California Control Consortium, and for the industrial data provided by the Eastman Chemical Company.
Keywords
- Granger causality
- Oscillation propagation
- Plant-wide oscillation
- Root-cause diagnosis
- Spectral Granger causality