Positive Time-Frequency Distributions: Development and Applications.

Abstract

Results of our research into the development and application of the positive (Cohen-Posch) time-frequency distributions (TFDs) are presented. These include (1) the development of a new computationally efficient weighted least squares method of implementation; (2) the development of a method to design informative priors for constructing Cohen-Posch TFDs; (3) studies of the relationship between instantaneous frequency and the conditional mean frequency of the positive TFDs, and the definition of the average frequency at each time; (4) a TFD based method of classification for nonstationary processes; (5) a new multi-window time-varying spectral analysis method; and (6) applications of the Cohen-Posch TFDs and conditional moments to machine vibration analysis and transient characterization. Copies of relevant publications are attached to the report.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1997
Accession Number
ADA328843

Entities

People

  • Patrick J. Loughlin

Organizations

  • University of Pittsburgh

Tags

Communities of Interest

  • Biomedical
  • Energy and Power Technologies
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Acoustic Scattering
  • Algorithms
  • Bandwidth
  • Classification
  • Computing-Related Activities
  • Data Science
  • Electrical Engineering
  • Frequency
  • High Resolution
  • Information Science
  • Joints
  • Machine Learning
  • Machines
  • Pattern Recognition
  • Signal Processing
  • Surface Warfare
  • Vibration

Fields of Study

  • Engineering

Readers

  • Control Systems Engineering.
  • Statistical inference.