The paper presents a movement strategy for Connected and Automated Vehicles (CAVs) in a lane-free traffic environment with vehicle nudging by use of an optimal control approach. State-dependent constraints on control inputs are considered to ensure that the vehicle moves within the road boundaries and to prevent collisions. An objective function, comprising various weighted sub-objectives, is designed, whose minimization leads to vehicle advancement at the desired speed, when possible, while avoiding obstacles. A nonlinear optimal control problem (OCP) is formulated for the minimization of the objective function subject to constraints for each vehicle. A computationally efficient Feasible Direction Algorithm (FDA) is called, on event-triggered basis, to compute in real-time the numerical solution for finite time-horizons within a Model Predictive Control (MPC) framework. The approach is applied to each vehicle on the road, while running simulations on a lane-free ring-road, for a wide range of vehicle densities and different types of vehicles. From the simulations, which create myriads of driving episodes for each involved vehicle, it is observed that the proposed approach is highly efficient in delivering safe, comfortable and efficient vehicle trajectories, as well as high traffic flow outcomes. The approach is under investigation for further use in various lane-free road infrastructures for CAV traffic.

Lane-free traffic is a novel research domain, in which vehicles no longer adhere to the notion of lanes, and consider the whole lateral space within the road boundaries. This constitutes an entirely different problem domain for autonomous driving compared to lane-based traffic, as there is no leader vehicle or lane-changing operation. Therefore, the observations of the vehicles need to properly accommodate the lane-free environment without carrying over bias from lane-based approaches. The recent successes of deep reinforcement learning (DRL) for lane-based approaches, along with emerging work for lane-free traffic environments, render DRL for lane-free traffic an interesting endeavor to investigate. In this paper, we provide an extensive look at the DRL formulation, focusing on the reward function of a lane-free autonomous driving agent. Our main interest is designing an effective reward function, as the reward model is crucial in determining the overall efficiency of the resulting policy. Specifically, we construct different components of reward functions tied to the environment at various levels of information. Then, we combine and collate the aforementioned components, and focus on attaining a reward function that results in a policy that manages to both reduce the collisions among vehicles and address their requirement of maintaining a desired speed. Additionally, we employ two popular DRL algorithms—namely, deep Q-networks (enhanced with some commonly used extensions), and deep deterministic policy gradient (DDPG), which results in better policies. Our experiments provide a thorough investigative study on the effectiveness of different combinations among the various reward components we propose, and confirm that our DRL-employing autonomous vehicle is able to gradually learn effective policies in environments with varying levels of difficulty, especially when all of the proposed rewards components are properly combined.

Controlling automated vehicles on large lane-free roundabouts is challenging because of the geometrical complexity and frequent conflicts among entering, rotating, and exiting vehicles. This paper proposes a comprehensive methodology to control the vehicles within the roundabout and the connected road branches. The developed real-time vehicle movement strategy relies on offline-computed wide overlapping movement corridors, one for each Origin-Destination (OD) movement, which delineate the admissible movement zones of corresponding OD vehicles. Also, space-dependent desired orientations are determined by destination, so as to mitigate potential vehicle conflicts and reduce trip distance. A distributed (per vehicle) movement control strategy, using two nonlinear feedback controllers (NLFC), for circular and straight movements, respectively, is employed to navigate each vehicle within the respective OD corridor toward its destination, accounting for the desired orientation and avoiding collisions with other vehicles; while boundary controllers guarantee that the corridor boundaries will not be violated, and the exit will not be missed. As an overly complicated case study, we consider the famous roundabout of Place Charles de Gaulle in Paris, featuring a width of 38 m and comprising a dozen of bidirectional radial streets, hence a total of 144 ODs. The pertinence and effectiveness of the presented method is verified via microscopic simulation and evaluation of macroscopic data.

This paper presents some new developments related to TrafficFluid-Sim, a lane-free microscopic simulator that extends the SUMO simulation infrastructure to model lane-free traffic environments, allowing vehicles to be located at any lateral position, disregarding standard notions of car-following and lane-change maneuvers that are typically embedded within a (lanebased) simulator. A dynamic library has been designed for traffic monitoring and lane-free vehicle movement control, one that does not impose any inter-tool “communication” delays that standard practices with the TraCI module introduce; and enables the emulation of vehicleto-vehicle and vehicle-to-infrastructure communication. We first summarize the various core components that constitute our simulator, and then discuss the new capability to utilize the bicycle kinematic model, additionally to the usual double-integrator model, as a more realistic model of vehicle movement dynamics, particularly for a lane-free traffic environment. Finally, we developed the necessary components so that the bicycle model can alternatively be combined with the use of global coordinates for more realistic simulation in road networks with curvature, such as roundabouts.