Urban Traffic Animation

We present a new agent-based system for detailed traffic animation on urban arterial networks with diverse junctions like signalized crossing, merging and weaving areas. To control the vehicle's intelligent car-following strategy on urban road networks, we extend the popular follow-the-leader model to generate smoother accelerating actions. The drivers will adjust their actions according to the information stored in road and junction structures, such as merging into another lane or going through a junction. Since the driving actions are decided by parameters that describe the vehicle's capacity and driver's preference, the system can produce various driving behaviors. In addition, in order to imitate the driver's various lane-changing behaviors on urban networks, we define the parametric description for driver's lane-changing intention and evaluate the safety of the proposed entire lane-changing process. The vehicle will finally change its lane either freely or cooperatively according to the local traffic situation. The free lane change often happens in sparse traffic flow, usually aiming at overtaking. When the vehicle is in a mandatory lane-changing situation and cannot perform free lane change, it will choose to do cooperative lane change, which requires other vehicle's cooperation.The existing 3D traffic simulation methods are only suitable for limited road networks (without intersections and merging regions) and simple driving style. Our system overcomes these limitations and also includes a continuous lane-changing model based on the driver's intention and safety evaluation. In the final animation, the vehicle changes its lane according to both the driver's intention and the local traffic situation.