Modeling drivers' lateral movement behavior under weak-lane disciplined traffic conditions

Abstract

In recent years the use of microscopic traffic simulation tools are increasing worldwide for selecting the most effective transport planning scheme and evaluate its economic feasibility as it can be used in laboratories for testing effectiveness of candidate traffic improvement initiatives before their actual field implementation. So, now a microscopic traffic simulator considering the traffic of Dhaka is a basic need for the transportation planner here. Development of the lateral movement model regarding the Dhaka traffic which is actually known as lane changing model worldwide carries a major importance as one of the base of the simulator. The lane changing is an important maneuver which defines the lateral movement on the road. As an effective component of the microscopic traffic simulator, a number of lane changing models have been developed so far in all over the world. But the models have been developed mainly for the homogeneous lane based traffic. The research has focused on developing a lateral movement model considering the weak lane based heterogeneous traffic of Dhaka where the vehicle can occupy any part of the road without strictly following the lane marking. The lateral movement model proposed in this research uses an econometrics based approach bearing in mind that the road consists of several number of small strips rather than lane. On the basis of the movement characteristics, Discretionary Lateral Movement Model and Mandatory Lateral Movement Model have been formulated then. The data used for developing these models have been collected from two locations of Dhaka City; Kalabagan and Shukrabad to represent two types of movement respectively using video cameras mounted on over-bridges. GPS equipped vehicles were run as well in a pre-specified route including these locations in order to supplement the video data. The recorded videos have been processed using software ‘TRAZER’ which provides vehicle trajectory data. The data have been fed into ‘MATLAB’ code then and the input variables for the model have been generated. Ultimately, two best models on the basis of physical and statistical significance for two different types of lateral movement have been considered as the final models out of various generated models. All the models have been generated using the statistical estimation software GAUSS. At last, the final models have been verified with different sets of trajectory data obtained from the same locations at different time. As the final models have been selected taking into account the most influential variables in case of lateral movement on the road and found significant, these models could be an effective components for the formation of the microscopic traffic simulator for the developing countries like Bangladesh.