Abstract
Road traffic crashes threaten thousands of drivers every day and significant efforts have been put forth to reduce the number of traffic crashes and their impact. Traffic congestion could be both a result of and a contributing factor to traffic crashes. The aim of this study is to investigate spatiotemporal traffic congestion and crash patterns to gain a better understanding of the causation of congestion and accidents, and their interaction. The Interstate 285 (I-285) in Georgia was used as a case study. With the aid of Geographic Information Systems (GIS), spatial clustering and densities of accidents were performed by following Anselin Local Moran’s method of spatial autocorrelation. The results indicated that the location of high-high accident clusters was in the northern half of the I-285 for all crash types. Additionally, geometric and traffic-related variables were correlated with accidents using logistic regression. The results showed that road segments involving merging, diverging, or weaving lanes had a positive correlation with the number of accidents. Specifically, the merging segments exhibited the highest crash frequency, followed by weaving and diverging segments. On the other hand, the road curvature did not play a significant role in crash occurrence, which is likely due to the gentleness of the road curvatures along the I-285 loop. However, the impact of acceleration on crash frequency remained inconclusive. It appeared that a lower average traffic speed correlated with a higher crash frequency, which may be due to a slow-down condition prior to crash occurrence. Table of Contents Abstract.......................................................................................................................................... ii List of Figures................................................................................................................................ 3 List of Tables.................................................................................................................................. 5 Introduction................................................................................................................................... 6 Objective ................................................................................................................................. 8 Literature Review ......................................................................................................................... 9 Overview of factors................................................................................................................. 9 Statistical Methods................................................................................................................ 14 Data Acquisition and Processing ............................................................................................... 15 Analysis........................................................................................................................................ 35 Local Moran’s I..................................................................................................................... 35 Logit Regression ................................................................................................................... 36 Results.......................................................................................................................................... 39 Conclusions and Recommendations.......................................................................................... 44 Research Limitations ................................................................................................................ 45 References.................................................................................................................................... 46 List of Figures Figure 1. An example of the XML file of traffic data................................................................... 16 Figure 2. Initial image created from collected speed data from I-285 cameras on March 28, 2018. ....................................................................................................................................................... 18 Figure 3. Average speed from October 1, 2017 to June 25, 2018 on Monday and Tuesday......... 19 Figure 4. Average speed from October 1, 2017 to June 25, 2018 for Wednesday through Friday. ....................................................................................................................................................... 20 Figure 5. Average speed from October 1, 2017 to June 25, 2018 separated by Saturday and Sunday........................................................................................................................................... 21 Figure 6. The averaging and denoising of Wednesday traffic....................................................... 22 Figure 7. Averaged weekday speed map with identified congestion. ........................................... 23 Figure 8. Location of congestion clouds for Clockwise traffic..................................................... 24 Figure 9 Location for congestion clouds for Counter-clockwise traffic....................................... 24 Figure 10. A visualization of the 42 traffic accidents (in red, size increased for clarity) on a denoised cloud map representing speed of traffic on March 28, 2018. ........................................ 25 Figure 11. All Monday accidents from 2016 to 2018 on I-285 (left) and averaged Monday I-285 traffic speed map (right)................................................................................................................ 26 Figure 12. Acceleration plot of March 28, 2018 with image in scale to show relation of crashes (red pixels) .................................................................................................................................... 27 Figure 13. Distribution of I-285 Crashes from 2016 to 2018 by collision category..................... 29 Figure 14. Heat map of Crashes on I-285 in 2018 by Collision type 1-4..................................... 30 Figure 15. Heat map of Crashes on I-285 in 2018 by Collision type 5 and 6.............................. 31 Figure 16. Map of I-285 accident data with closeup of interchange. ........................................... 32 Figure 17. Classification of road segment based on position in interchange................................ 33 Figure 18. Bar graphs of number of crashes by crash type and road segment type...................... 34 Figure 19. Road segment location of high-high clusters of crashes highlighted in blue. ............. 41 List of Tables Table 1. Descriptive statistics of variables.................................................................................... 37 Table 2. High-high clusters of accidents....................................................................................... 40 Table 3. Logit Regression Results for all crashes......................................................................... 42
Showing the abstract — retrieve the full paper via the Exa API.