Modeling the Effect of a Road Construction Project on Travel Time at Link-Level
Analytics
142 views ◎107 downloads ⇓
Abstract
Drivers observe variation in travel time due to congestion and delay on existing transportation facilities. A road construction project has a significant effect on travel time, leading to increased congestion, delay, and driver frustration. The effect on travel time due to the road construction project decreases as the distance from the road construction project location increases. The effect of the road construction project on travel time also extends to other roads that are connected to the road with the construction project. It also depends on traffic condition and time-of-the-day on these roads. Therefore, one needs to consider spatial dependency, and the influence on links within the proximity of the road construction project, over time, to model the effect of the road construction project on travel time at link-level. Findings from such research will help proactively plan construction activities on roads to mitigate mobility and congestion problems. Therefore, the goal of this research is to model the effect of a road construction project on travel time at link-level and help improve mobility of people and goods through dissemination or implementation of proactive solutions. The objectives of this research are 1) to examine travel time and travel time variations before, and during the road construction project period on a selected road, 2) to examine travel time and travel time variations on roads connected to the selected road with the construction project, 3) to examine the effect of network characteristics and develop models to estimate travel time on the selected road with the construction project and other connected roads, 4) to examine and develop models to estimate travel time by the time-of-the-day and the day-of-the-week during the road construction project period, and, 5) to compare the models to estimate travel time during the construction project period with models to estimate travel time before the construction project period.Data, from the year 2011 to the year 2016, was gathered from the Traffic Incident Management Systems (TIMS) and local agencies to identify a resurfacing construction project period for modeling the effect of road construction project on freeways and connecting arterial street links. The data obtained was processed by the time-of-the-day and the day-of-the-week to compute travel time performance measures using the Microsoft SQL 2012 software. A statistical t-test was conducted to examine the relationship between the change in travel time before and during the construction project period.The Statistical Package for the Social Sciences (SPSS) was used to check if the travel time data followed a normal distribution. From the normality test results, it was observed that the data followed a normal distribution. Generalized Linear Models (GLM) were developed with the average travel time on a link in the vicinity of the road construction project as the dependent variable. The characteristics of each link, such as the volume/capacity (V/C), the number of lanes, the speed limit, the shoulder width, the lane width, and whether the link is divided or undivided were considered as predictor variables for modeling. The characteristics such as the upstream and downstream link length, the upstream and downstream V/C, the upstream and downstream number of lanes, and, the upstream and downstream speed limit were also considered as predictor variables for modeling. Further, the time-of-the-day, the day-of-the-week, and the distance of the link from the road construction project were considered as predictor variables for modeling. The goodness-of-fit was assessed using the quasi likelihood under the independence model criterion (QIC) and the corrected quasi likelihood under the independence model criterion (QICC). The developed models were than validated using randomly selected samples for the same construction project. The samples used for validation were not used for model development.The travel times before and during the construction project period are significantly different than estimates obtained using the Bureau of Public Roads (BPR) travel time equation. A decrease in travel time was observed during the construction project period on the freeway links when compared to the before construction project period. Contrarily, an increase in travel time was observed during the construction project period on the connecting arterial street links when compared to the before construction project period.The results obtained indicate that predictor variables such as, the V/C, the upstream number of lanes, the upstream speed limit, the downstream V/C, and the downstream number of lanes have a significant effect on travel time before the road construction project period on freeway links. However, the V/C, the upstream link length, the upstream V/C, the downstream link length, the downstream V/C, and the downstream number of lanes have a significant effect on travel time during the road construction project period on freeway links. The V/C, the speed limit, the upstream V/C, and the upstream number of lanes have a significant effect on travel time before the construction project period on the connecting arterial street links. Similarly, the V/C, the speed limit, the upstream V/C, the upstream link length, the upstream number of lanes, and the downstream number of lanes have a significant effect on travel time during the construction project period on the connecting arterial street links.