CONTACT US: Contact info
|
Citation |
West LB, Heimbach CL. Highw. Res. Rec. 1973; 461: 17-29. |
|
Copyright |
(Copyright © 1973, National Research Council (U.S.A.), Highway Research Board) |
|
DOI |
unavailable |
PMID |
unavailable |
|
Abstract |
An analysis procedure is presented for determining the significant changes in instrumented car responses on adjacent short sections of a highway. The sections on which significant responses occur are then correlated by and with highway geometry elements, intersections and grades, and the number of accidents per short section. The procedure is demonstrated by the use of instrumented car data taken on us-70, a four-lane, divided, non-access-controlled highway in wake county, north carolina. The first step in the procedure consisted of removing systematic errors in location. Values were calculated by a least squares process for each driver response as measured by the instruments in the car for a specified short section of roadway. These values of the driver's response were compared sequentially for significant change by using a f-test. Those sections in which significant change was found were correlated on a binary basis (i.E., a zero given for the section without significance and a one given for sections with a significant response) with the intersections, grades of more than 4 percent, and accidents. Correlation coefficients were calculated for six different lengths of section. The highest correlations were found between accidents and those significant change sections whose instrumented car responses were speed change, running time, and total time. Significant driver responses did not correlate highly with intersections or grades. /author/ Language: en |
|
Keywords |
TRAFFIC SURVEYS; HIGHWAY ACCIDENTS |