Skip to main content
SHARE
Publication

Hidden Markov Modeling for Weigh-In-Motion Estimation...

by Robert K Abercrombie, Erik M Ferragut, Shane D Boone
Publication Type
Conference Paper
Book Title
Proceedings of the International Conference in Weigh-In-Motion
Publication Date
Page Numbers
13 to 21
Volume
N/A
Publisher Location
London/Hoboken, United Kingdom
Conference Name
6th International Conference on Weigh-In-Motion - ICWIM6
Conference Location
Dallas, Texas, United States of America
Conference Sponsor
International Society for Weigh-In-Motion (ISWIM), Transportation Research Board, and Texas DOT
Conference Date
-

This paper describes a hidden Markov model to assist in the weight measurement error that arises from complex vehicle oscillations of a system of discrete masses. Present reduction of oscillations is by a smooth, flat, level approach and constant, slow speed in a straight line. The model uses this inherent variability to assist in determining the true total weight and individual axle weights of a vehicle. The weight distribution dynamics of a generic moving vehicle were simulated. The model estimation converged to within 1% of the true mass for simulated data. The computational demands of this method, while much greater than simple averages, took only seconds to run on a desktop computer.