Search

Browse

• Communities
• Authors
• Year
• Titles
• Subjects

My Space

• Login

Motorway traffic control with connected automated vehicles

Markantonakis Vasileios

Full record

---

Summary

The Connected Automated Vehicle (CAV) revolution has begun, offering new benefits and mobility solutions to the society. The ability of CAVs to collect multiple types of data from the infrastructure and also exchange information with other connected vehicles (CV), promises to revolutionise the individual mobility within the upcoming years. Vehicle manufacturers experience the transition of becoming providers of innovative mobility solutions (IMS), replacing their up to now role of just producing conventional vehicles. Such technological advances and solutions usually come in the form of Advanced Drivers Assistance Systems (ADAS), already included at the majority of vehicles produced nowadays. These systems may contribute to improve safety and increase traffic flow efficiency on motorways, so as to develop a less congested and environmental friendly transportation network for all road users.The development, testing and validation of traffic control strategies for ADAS systems, is an interesting topic that gathered momentum in the research and industrial community during the last decade. The development of these strategies sorely lies on the CAVs ability to react in a controlled manner according to the prevailing traffic conditions, utilizing their own sensing technology. These strategies may suggest a new speed limit, a lane change maneuvre, a new time-gap value or a dedicated lane advice that needs to be applied by CAVs. In the current work, a series of local level traffic control strategies for mixed traffic are evaluated through microscopic simulations, on two real calibrated test-beds utilizing different penetration rates (PR) of CAVs, that act both as sensors and as actuators. The control strategies receive in real-time density, flow or speed measurements and deliver a series of actions to be enabled with the use of CAVs according to the policy adopted. Taking into account their simplicity in terms of control design, the control strategies showcase a satisfactory control behaviour to all the control scenarios examined.Analysing and optimising a multidestination network is typically considered a highly complex task. One destination may have more than one alternative routes, which are typically longer in distance and time. In most cases, road users select their route (i.e. which corresponds to the shortest route) based on travel time, in order to reach their destination as early as possible. This implies that during peak hours, the capacity of the route may attain or exceed its maximum value, leading to congested traffic conditions. Numerous studies in the literature address the problem of rerouting vehicles in order to prevent the formation of congestion at both urban and motorway environments, by proposing dynamic traffic assignment strategies using feedback strategies or optimal control problem formulations. In the current work, a first-order multi-lane macroscopic traffic flow model for motorway networks, incorporating capacity drop features, is utilized within an optimal control problem, to demonstrate the integrated use of various traffic control measures. Among these measures are mainstream traffic flow control, lane change control, ramp metering control and dynamic traffic assignment control. The optimal control solutions derived from the Quadratic Programming (QP) problem formulation are cast in a Model Predictive Control (MPC) framework for testing via application of the control actions for various PR of CAVs, within a microsimulation environment. A real motorway network with multiple destinations and routes was identified as the system for the needs of the microscopic evaluations. The results proved beneficial at all PR examined reaching the main goal of each one of the traffic control strategies involved. The lane changing actions offered a twofold contribution; that is to achieve an appropriate distribution of vehicles among all lanes and redirect the traffic using all the alternative routes. Furthermore, even at low PR, CAVs were able to drive the average speed of all vehicles close to the speed limit ordered; thus, applying MTFC actions on the mainstream motorway. Moreover, the use of RM actions at specific on-ramp merges were necessary to keep the merging area at an under critical density regime and avoid the formation of congestion on the motorway corridor.

Available Files

Services