Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Machine failure forewarning via phase-space dissimilarity measures.

L M Hively1, V A Protopopescu

  • 1Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USA.

Chaos (Woodbury, N.Y.)
|June 11, 2004
PubMed
Summary

This study introduces a new data-driven method to detect changes in complex systems, improving early machine failure warnings. The approach uses phase-space dissimilarity for more consistent and powerful fault detection than traditional methods.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Detecting dynamical changes in time series using the permutation entropy.

Physical review. E, Statistical, nonlinear, and soft matter physics·2004
Same author

Epileptic event forewarning from scalp EEG.

Journal of clinical neurophysiology : official publication of the American Electroencephalographic Society·2001
See all related articles

Area of Science:

  • Nonlinear dynamics
  • Complex systems analysis
  • Predictive maintenance

Background:

  • Traditional methods for analyzing nonlinear systems and predicting machine failures often lack consistency and discriminating power.
  • Identifying dynamical changes in complex systems is crucial for effective prognostics and health management.

Purpose of the Study:

  • To develop a model-independent, data-driven approach for quantifying dynamical changes in nonlinear processes.
  • To apply this method for reliable machine failure forewarning using vibration power data.

Main Methods:

  • Utilized time-delay phase-space reconstruction on time-windowed data to derive discrete invariant distribution functions.
  • Quantified system condition changes using dissimilarity measures between test case and baseline distribution functions.

Related Experiment Videos

  • Analyzed time-serial mechanical vibration power data from motor-driven systems with induced faults.

    • Main Results:

    • The proposed phase-space dissimilarity measures demonstrated superior consistency and discriminating power compared to conventional statistical and nonlinear measures.
    • The method effectively identified condition changes indicative of impending machine failures.
    • The approach proved robust in analyzing data from systems with accelerated failures and seeded faults.

    Conclusions:

    • The developed phase-space dissimilarity approach offers a more reliable and consistent method for detecting dynamical changes in nonlinear systems.
    • This technique shows significant promise for timely and accurate forewarning of equipment failure in industrial applications.
    • The model-independent nature of the method allows for broad applicability across various nonlinear dynamical systems.