Congestion Identification for Multilane Highway by Vehicle Lane Change as Driver Behavior

Abstract

This work introduces a new way to assess the condition of highways and classify them as congested or uncongested highways. This was done by recording a video and sorting the number of lane change orders instead of finding the traffic density, delay, and flow rate. Then extracting data from the video and performing a microscopic simulation for it to study the driver's behavior and create the statistical model. Thus, obtaining the mathematical relationships and coefficients that illustrate that process, after verifying the reliability and calibration of these variables. This research aims to facilitate the highway assessment process, as an alternative or supplementary method of highway evaluation by the Highway Capacity Manual 2010. It is adequate to record a video and count the number of vehicles that change their lanes and then compare them with the results that have been determined in the search to judge the highway is congested or not. The study area was achieved by choosing five segments of multilane arterial roads (six lanes divided) on the left coast of Mosul city. These segments were selected to be as similar as possible in the geometric configuration as well as like traffic flow. After that, cameras were installed according to criteria to record traffic movement and for a time not less than ten hours for each segment, and on the workdays for traffic movement. Then, the videos were uploaded using the Goodvision traffic data recognition program. These outputs were converted to be entered into the PTV-VISSIM traffic simulation program in addition to the other geometric information needed by the program to complete the analysis process and find traffic flow parameters. In the extraction process, the normal distribution and the uniform nature of the data representation of the observed event on the segments were examined using the Z-test and the K-S test, the results were normal and significant. Then the results were verified and the traffic parameters such as density and flow rate were checked to ensure the nature of the traffic flow. The process was also verified using the vehicle motion graphic representation method (Vehicle Trajectory), which is the initial examination to estimate the new variable. A new variable has been extracted, which is one of the traffic movement variables, which depends on the percentage of complete and proper lane change completion within the parameters imposed by the simulation program, it’s the Lane Change Occupancy factor (LnChOc.). Then the statistical model was derived using SPSS for the new variable. It describes traffic conditions on highways and the results are verified and calibrated in more than one way. It’s a simple statistical model was obtained to estimate the traffic movement from this variable and vice versa from the following equation:, Which gave results for critical traffic on multi-lane roads in the Mosul city as follows: - Ø Ø Ø