Date of Award

7-2004

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Electrical Engineering and Computer Science

First Advisor

Barry G. Grossman

Second Advisor

Frank Caimi

Third Advisor

Syed H. Hurshed

Fourth Advisor

John J. Engblom

Abstract

This report studies the effect on the accuracy o f weigh-in-motion (WIM) measurements due to the non-uniformity along the length o f a fiber-optic WIM sensor, and demonstrates an effective compensation method. Accurate WIM calculations require accurate input signals from the road sensors. Today’s WIM systems that use piezzo-electric sensors, electronic strain gages or load cells have errors anywhere from +/-6% to +/-50%, with -i-/-15-20% being the typical range. Fiber-optic WIM sensors have promising advantages. However, the state of the technology is still in its infancy consisting mostly of qualitative studies. This report isolates the problem of non-uniform sensitivity in a microbend fiber-optic sensor from other causes such as road surface conditions, mechanical vibrations, and wind gusts which also cause inaccurate WIM measurements. The hypothesis is that the sensor is significantly non-uniform along its length which introduces and contributes to errors in the weight calculations, and that by fully characterizing the sensor and devising a way to measure the location of the tire impact and the tire width through the use of a new sensor configuration, such nonuniformity can be significantly compensated. Furthermore, the same tire location and width data can be used to reduce the classification error rate, which also leads to additional increase in WIM accuracy, especially in portable WIM systems. Taking into consideration of non-uniformity in a fiber-optic WIM sensor to improve the accuracy of WIM measurement is a new idea, and the demonstrated technique could be used to compensate not only the non-uniformity of the fiberoptic sensor in a laboratory, but it could also be used to compensate for post-installation non-uniformities caused by such things as materials used to construct the road and road surface contour. Because less uniform sensors could be compensated, it brings the possibility of improving the accuracy of the WIM system and reducing cost at the same time. Here, a fiber-optic sensor with +/-20% tolerance was compensated to +/-5~6% and mis-classification of vehicles in heavy traffic was reduced.

Comments

Copyright held by author

Download

Share

COinS