Transportation Research Interdisciplinary Perspectives (May 2023)
Balancing the efficiency and robustness of traffic operations in signal-free networks
Abstract
Integration of artificial intelligence and wireless communication technologies in Connected Automated Vehicles (CAVs) enables coordinating the movement of the platoons of CAVs at signal-free intersections. The efficiency of the platoon coordination process can be improved by reducing the spacing between successive platoons to increase capacity; however, such improvement in efficiency can have adverse impacts on the robustness of the coordination process. In this research, we balance the trade-off between the efficiency and robustness of traffic operations in signal-free networks at a macroscopic scale. To this end, we use a rule-based approach to express the process of coordinating CAV platoons at intersections as a set of governing equations that provide an analytical basis to develop a stochastic model for traffic operations. We derive the platoon synchronization success probability for a general distribution of the error in synchronizing the movement of platoons in crossing directions and formulate the expected capacity as a function of the synchronization success probability. We then balance the trade-off between efficiency and robustness at a macroscopic scale by adjusting the average spacing set between successive platoons. In urban networks, adjusting the spacing between successive platoons also changes the vehicular density and consequently the traffic speed. We account for the interrelationship between the traffic speed and inter-platoon spacing in balancing the trade-off between the efficiency and robustness of traffic operations using the concept of the Macroscopic Fundamental Diagram (MFD) and extend the stochastic traffic model to the network level. We evaluate the analytical results of the research using a simulation model. The numerical results of the research show that optimizing the system by adjusting the platoon spacing can improve robustness by 13% at the cost of a 4% reduction from the maximum capacity at the network level.
