Abstract
The situation in which a vehicle has to avoid a collision with an obstacle can be difficult to realise in optimum conditions when the roads are crowded. This paper uses the advantages of vehicle grouping and vehicle-to-vehicle (V2V) communication, and proposes a control architecture, which ensures a safe merging between the vehicles from two platoons. The architecture is formed by three layers, with the following tasks: i) to analyse the environment and to decide the best action for a certain vehicle, ii) to plan the new trajectory, and iii) to follow it at an imposed velocity or distance to the vehicle in front. The vehicles are equipped with a trajectory planner designed using two methods: the first one is based on a polynomial equation, and the second one is based on the model predictive control (MPC) algorithm. Each vehicle is also equipped with a trajectory follower, which has a cooperative adaptive cruise control (CACC) functionality based on a distributed model predictive control (DMPC) formulation. Also, the paper proposes a solution to compensate the data-packet-dropouts that are induced by the wireless communication network used to exchange information between vehicles. Moreover, to accommodate various realistic scenarios in the same control framework, the cost function for the DMPC algorithm was designed to take into account different communication topologies. The proposed architecture was tested in a simulation scenario, in which two platoons have to merge in order to avoid a fixed obstacle and the results show its efficiency.
Highlights
The current studies in the field of autonomous vehicles have the following main directions: i) to ensure people’s safety by avoiding collisions, ii) to reduce costs and pollution by decreasing fuel consumption, iii) to improve traffic condition by avoiding congestions, and iv) to optimize the space occupied by vehicles by maintaining a small safety distance between them [1]
Using vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication the vehicles can make use of more information about their neighbours and environment. This information is afterward used to improved either: i) the performance in traffic or ii) the performance of the existing solution (e.g., the ACC algorithm is turned into a cooperative adaptive cruise control (CACC) algorithm, in which a follower vehicle benefits by receiving the information regarding the velocity of the vehicle in front [11]–[14] )
This paper proposes a control architecture formed by three levels, which can be used by the vehicles driving in platoons to ensure a safe travelling
Summary
The current studies in the field of autonomous vehicles have the following main directions: i) to ensure people’s safety by avoiding collisions, ii) to reduce costs and pollution by decreasing fuel consumption, iii) to improve traffic condition by avoiding congestions, and iv) to optimize the space occupied by vehicles by maintaining a small safety distance between them [1]. Applications of merging vehicles receive a great interest nowadays These studies mainly use the point model for the vehicles and the control solutions are based on LQR or MPC algorithms [26], [33], [38]–[40]. The MPC algorithm uses a simple model for the dynamics of the vehicle to estimate its position and to minimise a cost function in order to lead the vehicle from an initial position to a given target position; Level III is composed of a CACC algorithm and a trajectory follower: the CACC is implemented using a DMPC algorithm and its task is to ensure that the vehicles from a platoon are travelling at the imposed velocity and maintain a safe distance between them. The symbol ||Xi||∞ denotes the infinity norm and represents the maximum of Xi ∈ Rn
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
