Abstract

Modeling is a safe and efficient approach to solving real vehicle traffic problems. It cooperates with/without simulation to provide feasible methods of analysis, observation, and verification. Modeling can give significant insight into complex systems. Modeling Intelligent Transportation Systems (ITSs) represents a crucial challenge in planning and controlling vehicle traffic congestion. Predicting the vehicle’s flow states on roads is the most important challenge in transportation systems. Numerous car-following models have been proposed and developed to illustrate the behavior of moving vehicles. These models are based on real driving assumptions, taking into account velocity and acceleration parameters for each vehicle. The application of car-following models represents an important research direction in enhancing ITSs. In this paper, a car-following model is implemented using a specific vehicle traffic dataset (highD) to predict the vehicle’s next velocities after a succeeding period of time based on regression fundamentals. The vehicle’s Velocities are based on the driver’s behavior. The driver can accelerate, decelerate, or maintain the same speed during any time period of the vehicle journey. Certain threshold values are created by analyzing the recorded real dataset to be used in predicting the next acceleration value for each vehicle at each time period. A regression curve is proposed for each vehicle. From the proposed curves, equations are created to represent the vehicle’s velocity. These created equations can be used to predict the vehicle velocity at any given time mathematically.