United States Department of Transportation
i
Field evaluation of smart sensor vehicle detectors at railroad grade crossings - volume 4 : performance in adverse weather conditions.
-
2015-01-01
[PDF-1.04 MB]
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Corporate Publisher:
-
Abstract:The performance of a microwave radar system for vehicle detection at a railroad grade crossing with quadrant gates was evaluated in
adverse weather conditions: rain (light and torrential), snow (light and heavy), dense fog, and wind. The first part of this report compares
the results of the modified system setup in adverse weather conditions with those from good weather conditions (as presented in
Volume 3 of this study). Then, the results of a re-modified system setup were compared to the results for the modified system setup in
good and adverse weather conditions. The re-modification was in response to increased detection errors in adverse weather conditions.
With the modified setup, system performance was sensitive to the adverse weather conditions. In torrential rain, false calls increased to
24.82%–27.08% (e.g., May 28 and June 1) when there was some traffic on the crossing. However, when there was torrential rain but
only one vehicle (e.g., May 31) or no traffic flow (e.g., June 10), the radar units generated 15 false calls on each of those 2 days. For all
heavy snow datasets combined, missed calls by a single radar unit and by the two radar units working as a combined unit (i.e.,
systemwide) represented 13.51% and 11.66% of the loop calls, respectively. The most severe snow effects were found during freezing
rain/ice. In dense fog, false calls increased to 11.58%, and all false calls were generated when the gates were moving or in the down
position. Wind did not affect system performance, and the errors were similar to those in good weather conditions. With the re-modified
setup, the frequency of errors in heavy rain and heavy snow conditions was reduced and system performance was similar to the good
weather, light rain, and light snow conditions. In heavy rain, false calls in the re-modified setup were reduced to 2.6% compared with
30.5% in the modified setup. This reduction was the result of a significant decrease in the false calls generated without objects in the
crossing. The re-modified setup eliminated the systemwide missed calls in heavy snow. The re-modified setup also reduced the false
calls to less than 1% in good weather, light rain, and light snow conditions and practically had no missed, stuck-on, or dropped calls.
Results indicate that re-modifications improved the performance of detection system.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
