Trajectory Optimization Through Connected Cooperative Control For Multiple-Vehicle Flocking

The ever-increasing number of vehicles that use the current infrastructure brings many disadvantages, among which one can emphasize traffic congestion and decreased safety, mostly due to human error. While the human drivers are aided more and more by different automated functionalities that help them make the best decisions at certain times or may even be replaced by higher-level functionalities, automated vehicles are far from being deployed in series on the roads because it is difficult to ensure functional safety at all times. Moreover, automated vehicles would solve only half of the problem, i.e., removing the human factor from the loop, but the number of vehicles would not decrease which leads to the same traffic congestion. One solution would be to reduce the space between the vehicles, but this implies even higher functional safety of the automated features. Another idea is to make use of the vehicle platooning concept based on cooperative adaptive cruise control and extend it to the multiple-lane use-case. This would involve coordinating a group of vehicles traveling on a multiple-lane road to maintain small gaps between them even at highway speeds. The solution implies an exchange of information between vehicles, such that their trajectories are optimized at all times. Thus, this paper considers the design and development of a control architecture for multiple-lane vehicle flocking based on their cooperative decisions. The simulation results obtained using a simulator based on SUMO and Matlab illustrate the capabilities of the proposed methodology.