Non-Gaussian and persistence measures for control loop quality assessment
AbstractThis paper presents review and comparison of alternative methodologies for control performance assessment. The approach uses nonlinear time series analysis, such as non-Gaussian statistics, fractal, crossover analysis, or entropy-based approaches. There is a presented practical rationale for the analysis. Evaluation is based on the real data gathered from industrial systems. Non-Gaussian analysis starts with statistical methods using different probabilistic distribution functions. As another potential measure, the Hurst exponent is calculated using different approaches. Finally, R/S plot analysis together with crossover point phenomenon discussion is presented. The paper ends with conclusions and presentation of open issues attractive for further development. Fractal analysis has already met general acceptation in several areas of research. Starting from the analysis of economical data, it was successfully used in different areas of science, such as physics, biology, medicine, computer graphics, network traffic analysis, telecommunication, and time series prediction. On the other hand, there are still some areas of research, where fractal analysis is nonexistent. One of them is control theory and automation. However, the author observed that there are practical aspects of control engineering, in which standard linear and Gaussian analysis may not be appropriate and possibly other approaches should be proposed. It appears that in industrial control system, performance data have non-Gaussian properties and fractal analysis may improve assessment enabling deeper insight into underlying phenomena. This paper addresses those issues. At first, the research area is presented and then it is followed with the review and comparison of standard approaches with nonlinear fractal based measures. The analysis is illustrated with data examples from industrial sites. The paper concludes with results synthesis and presentation of open issues attractive for further research.
|Journal series||Chaos, ISSN 1054-1500|
|Publication size in sheets||193973.35|
|ASJC Classification||; ; ;|
|Project||Development of methodology of control, decision support and production management. Project leader: Zieliński Cezary,
, Phone: 5102, start date 19-05-2015, end date 31-12-2016, 504/02233/1031, Completed
|Score|| = 40.0, 02-02-2020, ArticleFromJournal|
= 45.0, 02-02-2020, ArticleFromJournal
|Publication indicators||= 16; = 7; : 2016 = 1.023; : 2016 = 2.283 (2) - 2016=2.312 (5)|
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.