Methods and apparatus for automatically detecting differences between similar but different states in a nonlinear process monitor nonlinear data. Steps include: acquiring the data; digitizing the data; obtaining nonlinear measures of the data via chaotic time series analysis; obtaining time serial trends in the nonlinear measures; and determining by comparison whether differences between similar but different states are indicated.
The present invention relates to methods for analyzing nonlinear data
from diverse sources, including but not limited to electroencephalogram
and magnetoencephalogram brain wave data, electrocardiogram data, and
motor current data from processes
such as metal machining operations and electrically-driven pumps.
A deviation from normal process conditions can indicate performance degradation or the onset of imminent failure. Real- or near-real-time monitoring of the process condition can detect this deviation in data from current, voltage, acceleration, acoustic emission, pressure, power, and other measurable quantities. Prompt response to a significant deviation can help to maintain optimal performance and can avert disastrous failure.
The theory of nonlinear dynamics provides a basis for understanding and potentially controlling many complex physical and engineering systems. For example, an extensive literature exists for nonlinear dynamics in the brain and related work as cited in U.S. patent application Ser. No. 08/619,024, U.S. patent application Ser. No. 08/619,030, and U.S. patent application Serial No. 08/619,031 hereinabove incorporated by reference. Other documents listed below refer to earlier work using linear and nonlinear methods of analysis. Numbers in parentheses within the text of the disclosure hereinafter refer to documents as listed below.
Tech Transfer Details
- http://inspire.ornl.gov/Document/View/d5a05ecc-27 ...
- Hively; Lee M.; Ng; Esmond G.
- Patent #5815413
- 3 : Analytical and experimental critical function and/or characteristic proof of concept
CategoriesSimulation & Modeling
Sharing and Social Links