This paper presents a reservation-based autonomous intersection control algorithm, named Autonomous Control of Urban TrAffic (ACUTA). The ACUTA allows centralized management of autonomous vehicles within a certain distance from the intersection to manage vehicles to pass the intersection with no conflict and few stops. To address some operational issues identified in previous studies on reservation-based autonomous intersection management, three operational improvement strategies are introduced and incorporated in the ACUTA. These strategies have been evaluated and proved to be effective under certain conditions. Along with the operational improvements of the ACUTA, its implementation in VISSIM is significant to achieve standardized evaluation of operational and safety performance using a widely applied simulation platform. This has never been accomplished in previous studies. Additionally, VISSIM simulations are established to compare the operational performances between a multi-tile ACUTA intersection and a signalized intersection and between a single-tile ACUTA intersection and a four-way stop intersection. Results obtained from both comparisons suggest that the ACUTA is more effective in maintaining a much lower intersection delay as well as a shorter or even empty queue. Sensitivity analyses are conducted on four configurable parameters of the ACUTA for a better understanding of how optimization can be done in the future. Granularity (number of tiles) is found critical in maintaining low intersection delays. Three other parameters are also found to have considerable impacts on the operational performance.
↧