Identification of Wheelset/Rail Creep Coefficients from Dynamic Response Data Using the Maximum Likelihood Parameter Identification Technique.

Abstract

This thesis explores the application of the maximum likelihood parameter identification technique to determine the wheel/rail creep coefficients using dynamic response data. The equations of motion for the dynamically scaled wheelset are presented and the reduced form of the maximum likelihood equations as applicable to the dynamically scaled wheelset model are developed. The maximum likelihood equations were formulated into a maximum likelihood algorithm which was implemented in Fortran IV. Using simulated wheelset data, the effects of a random input representation of the track versus a deterministic input with uncertainty representation are determined. The effects of various levels of measurement noise are also examined. This preliminary analysis indicates that the deterministic representation of the track input yields better results. Representing the track as a random track input requires further investigation into the effects of longer data records and smaller time steps on the performance of the maximum likelihood algorithm. (Author)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 1979
Accession Number
ADA107447

Entities

People

  • William N. Herzog

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Energy and Power Technologies
  • Space

DTIC Thesaurus Topics

  • Analog Computers
  • Calibration
  • Computer Programming
  • Computer Programs
  • Computers
  • Differential Equations
  • Digital Computers
  • Dynamic Response
  • Engineering
  • Equations
  • Equations Of Motion
  • Mathematical Filters
  • Measurement
  • Research Facilities
  • Riccati Equation
  • Sampling
  • Statistics

Fields of Study

  • Engineering

Readers

  • Robotics and Automation.
  • Statistical inference.
  • Structural Dynamics.