Daytime and Nighttime Passenger Car Operating Speed Models for Two-lane Rural Highways

Open Access
Author:
Kulis, Philip Nicholas
Graduate Program:
Civil Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
March 20, 2014
Committee Members:
  • Eric Todd Donnell, Thesis Advisor
Keywords:
  • Operating speed models
  • Nighttime speeds
  • Rural highways
  • Three-stage least squares
  • ordinary least squares
Abstract:
Operating speed models need to be developed to understand how traffic control and geometric features affect driver’s speed selection. Once operating speed models are developed, they can be incorporated into the geometric design process to ensure that roadways are designed to maintain consistency and anticipated operating speeds coincide with the design speed. While it is well known that driving at night poses a significant risk, it is not well understood how drivers behave at night. There is a significant body of research on passenger car operating speeds, which focuses on daytime and ideal conditions, but very little research has focused on nighttime operating speeds. To discover some of the potential differences between operating speeds during the daytime and nighttime, three-stage least squares (3SLS) regression was used to estimate both the speed magnitude (mean speed) and speed dispersion (speed deviation) during the daytime and nighttime. The models for estimating mean speed and speed deviation were then compared to each other to determine if there were differences, and the result of a transferability test shows that the two models are different. The comparison of the coefficients in the models and the transferability test show that there are differences between how drivers select their operating speeds during the daytime and nighttime, and these differences should be considered during the geometric design process. At higher approach speeds, there may be less of a concern about design consistency and more of a concern about vehicles exceeding the design speed during the nighttime since the models show that drivers maintain a speed closer to their approach speed through the curve more so during the night than day. Possible ways to lower operating speeds during the nighttime and have them more closely match operating speeds during the daytime are to reduce approach speeds since drivers maintain a more constant speed entering a curve during the nighttime than daytime or increase the delineation of the curve so drivers can judge the geometric layout of the curve better. This research can also be expanded upon to explore other potential differences between operating speeds during the day and night. Based on the additional analysis that is done and the coefficients of additional variables, it may be found that there are other variables that have a significant difference on operating speeds during the daytime and nighttime, and these differences should be considered in geometric design policy and processes.