The present paper proposes a collaborative control strategy for CAV movement in a lane-free environment, where CAVs have different priority values. The priority values are designed to introduce fairness, relying on the interaction history of every CAV. These priority values are the output of the Stationary Nash Equilibrium (SNE) of a modified Karma game. The modification is proposed to tackle a proportional resource allocation problem. The outcome of the modified Karma game is then used to determine the priority of CAVs in their interactions with the other vehicles in their neighborhood, which may constitute potential threats, so-called Threatening Vehicle Cluster (TVC). An MPC is designed at an upper level to enforce an overall fair collaboration among CAVs so as to avoid collisions. Finally, a simulation case study analyzes the proposed control approach regarding performance, efficiency, and fairness.

The concept of Internal Boundary Control (IBC) has been recently incorporated into Lane-free vehicle movement with the aim of maximizing the utilization of crossroad infrastructure in both directions. Although IBC is generally successful in alleviating congestion, there are scenarios where it may not be able to fully dissolve it. Employing Ramp Metering (RM) can be an effective strategy for managing the flow of traffic from the onramp to the mainstream in this situation. This paper proposes an online scheme optimization approach for integrated IBC and RM in lane-free traffic. For controller design, a dynamic traffic flow model based on a first-order macroscopic model with the driver's anticipation term is considered. The nonlinear and multivariable characteristics of this macroscopic highway model make it unlikely to be practical to create an online optimization mechanism for it, primarily due to the expected high computational expenses. So, a feedback linearization approach is employed to tackle the complexities of the macroscopic model. Following that, a linear Model Predictive Control (MPC) is designed based on the feedback linearized model to obtain an efficient and real-time solution. The suggested approach undergoes examination through a series of simulation tests to evaluate its performance. The simulation results verify the efficacy of the proposed approach in enabling real-time and efficient traffic management in future lane-free highway systems.

This paper aims to develop a traffic control approach that provides an optimal real-time solution for large-scale highway networks. The traffic flow model’s nonlinearity is the main reason for the complex optimization problems that consequently require high computational effort. Feedback linearization is a well-known approach in nonlinear systems control. It can provide an exact linear representation of the original nonlinear system. Thus, it facilitates further steps in the controller design. This research combines the feedback linearization method and linear MPC to simultaneously guarantee optimal performance and real-time feasibility. While developing feedback linearization for METANET model, we discovered a pattern that describes the expansion of the control signals (ramp metering and variable speed limits) and disturbances effects through the highway networks. Utilizing this pattern, we design a generic feedback linearization control law for METANET model. The control law provides the key connection between the linearized and original model. Finally, we complete the methodology by employing a linear MPC to regulate the linearized model. The performance of the designed method is evaluated by conducting comprehensive simulation studies, including a large-scale network. The simulation results are promising. Eventually, comparing the developed methodology with an equivalent nonlinear MPC verifies a substantial improvement in computational costs.