Citation

Himes SC, Donnell ET, Porter RJ. Transp. Res. Rec. 2011; 2262: 31-41.

Copyright

(Copyright © 2011, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

10.3141/2262-04

PMID

unavailable

Abstract

Operating speed prediction models along horizontal curves of two-lane rural highways have been the topic of a significant body of literature. Operating speed models for two-lane rural highway tangents and for horizontal curves on low-speed urban streets also exist but are less developed. These models, estimated primarily by means of ordinary least squares regression, generally use aggregate-level data to estimate the effect of geometric design variables on the mean or 85th percentile operating speed but do not explicitly consider measures of speed dispersion (e.g., standard deviation). Most existing models consider speeds of free-flow passenger vehicles; traffic flow effects on speed are generally not quantified. Finally, published literature shows that researchers have not converged on a consistent modeling practice about the use of posted speed limit as an explanatory variable in operating speed models. Collectively, these issues may lead to design consistency assessments that are not truly reflective of the operating conditions along a two-lane highway. These issues are explored in this paper with the use of operating speed data collected at multiple points along eight two-lane highway segments. A simultaneous equations approach is used to estimate a three-equation model of the posted speed limit, mean speed, and speed deviation. The predicted 85th percentile speed is compared with the observed 85th percentile speed along a high-speed, two-lane rural highway. Predicted 85th percentile operating speeds along the same high-speed highway determined with the Interactive Highway Safety Design Model design consistency module are also compared with observed 85th percentile speeds as a basis for evaluating the simultaneous equations approach. Finally, the paper proposes an alternative approach to evaluating design consistency that uses inferred design speed.