Advances in Uncertainty Treatment in FHWA Procedure for Estimating Annual Average Daily Traffic Volume

Abstract

This paper proposes an approach that estimates annual average daily traffic (AADT) of a road section from incomplete data. This approach improves the accuracy and interpretability of the results while preserving the concept of the current FHWA procedure. When the short-period traffic count (SPTC) of a road section is given, a road group with a known AADT and a similar traffic pattern is identified. The AADT of the road section in question is estimated by adjusting the AADT associated with the road group by the degree of similarity. The uncertainty associated with the similarity is measured by nonspecificity and discord. The model was tested with data obtained in the Province of Venice, Italy. The analysis considered the characteristics of SPTC, including duration and day of the week. Estimates obtained with the proposed method and two existing methods were compared. The proposed method was found to produce more accurate results than the previous methods. Weekday 48-h short counts were found to be the best sample SPTC for AADT estimation. Furthermore, the measures of uncertainty helped to interpret the quality of the estimates. SPTCs with a lower value of discord were found to yield better AADT estimates, and a high value of nonspecificity indicated uncertainty with respect to matching with the groups. These measures can also indicate the need for additional data collection.