Evaluation of driver Acceleration/Deceleration behavior at signalized intersections using vehicle trajectory data

Vehicle Acceleration/Deceleration (A/D) characteristics are fundamental measures often used in designing intersection geometrics, A/D lanes, analyzing vehicle fuel consumption and emission, and traffic simulation modeling. The flow scenario at a signal-controlled intersection involves stop-and-go conditions with variability of A/D characteristics. This study attempts to analyze the vehicle's A/D behavior at an urban signalized intersection under weak lane disciplined mixed traffic stream using vehicle trajectory data. The result shows that vehicles like car and motorized two-wheeler are associated with high acceleration behavior at lower speeds and need less time and distance to accelerate. Also, these vehicles can decelerate at a higher rate at a lower speed during the stop or clearing of the intersection. The statistical results show that a single and multi-regime polynomial model is more suitable to represent the behavior of A/D characteristics in terms of the approach speed of vehicles.