Date of Award
Master of Science
Yue Liu, Lingqian Hu
FREEVAL-RL, NPMRDS, Travel Time Index, Travel Time Reliability, TTI, VSPOC
Travel time reliability aims to quantify the variation of travel time by using the entire range of travel times for a given trip, for a selected time period over a selected horizon. A trip can occur over a segment, facility or any subset of the transportation network, for the purpose of calculating travel time reliability. As one of the most important performance measures, travel time reliability reports the number of trips that fail or succeed according to a predetermined standard. Unreliability is usually caused by the interaction of factors that influence travel times, such as fluctuations in demand due to daily or seasonal variation, or special events, traffic control devices, traffic incidents, inclement weather, work zones, and physical capacity. These factors collectively produce travel times that can be better presented by a probability distribution.
A well-accepted measure of travel time reliability is the Travel Time Index (TTI) formulated as the ratio of travel time in the peak period to the travel time at free-flow conditions. In this thesis, the Travel Time Index values were calculated and compared from two different kinds of data sources: probe vehicles and fixed location detectors. Speed from vehicle probe data can be retrieved from the National Performance Management Research Dataset (NPMRDS) and the freeway segment speed can be calculated by dividing the segment length by the total travel time. Spot speed from fixed location detectors can be retrieved from the Wisconsin’s Archived Data Management Systems (ADMS), V-SPOC (Volume, Speed and Occupancy) which measures the speed at certain locations of a segment. The free flow speed also varies by data source. In the V-SPOC data, the posted speed limit is considered to be the free flow speed and in the NPMRDS data, the reference speed which is the 85th percentile speed of all observed sample speeds is considered to be the free flow speed.
The effect of data quality on the TTI values is also examined in the thesis. Inductive loop detectors are a major source of traffic information, but they are often criticized for generating missing and faulty data which compromise real-time traffic control, operations, and management. There is no doubt that the quality of data will affect the accuracy of the calculation of Travel Time Index and its influence needs to be quantified. This study area was chosen to be the one that contains all different kinds road segments like basic, weaving, on ramp and off ramp segments. The result shows that the removal of invalid data improves the TTI index in the congested traffic conditions.
Lastly, a traffic simulation application, FREEVAL-RL tool, was applied to calculate the Travel Time Index. The sensitivity analysis of some important parameters used in the FREEVAL-RL Tool was performed. Calibration procedure was designed and carried out for the tool to reflect the real-world scenarios such as are Capacity Adjustment Factor, jam density and capacity drop. The outcome of the calibrated model was consistently matched to the travel time distribution in terms of mean, 50th percentile, 80th percentile, 95th percentile Travel Time Index (TTI) reported in the NPMRDS data.
Pankaj, Ambily, "A Comprehensive Study on the Estimation of Freeway Travel Time Index and the Effect of Traffic Data Quality" (2019). Theses and Dissertations. 2234.