Consensus-based local information coordination for the networked control of the autonomous intersection management

Abstract

Autonomous intersection management (AIM) will be a future method for improving traffic efficiency in the urban area. Instead of using the traffic signal control like nowadays, it uses wireless communication with autonomous vehicles to support the management of road traffic more safely and efficiently. A single AIM shows an exceptional performance in managing traffics at an intersection. However, it could not be represented a traffic in the real world, which is composed of multiple intersections. We show that coordination of traffic information among vehicles and infrastructures is an essential part of macroscopic traffic management. Coordination of traffic information among the network of AIMs is the key to improve the overall traffic flow throughout the network not only has an optimal flow in some intersections and very heavy traffic in others. In this paper, we introduce the distributed control to a graph-based intersection network to control traffic in a macroscopic level. Vehicle to infrastructure and infrastructure to infrastructure communication are used to exchange the traffic information between a single autonomous vehicle to the network of autonomous intersections. We implement a discrete time consensus algorithm to coordinate the traffic density of an intersection with its neighborhoods and determine the control policy to maximize a traffic throughput of each intersection as well as stabilizing the overall traffic in the network. We use the Greenshields traffic model to define the boundary condition of various traffic flows to the corresponded traffic density and velocity. Our proposed method represents the ability to maintain traffic flow rate of each intersection without having a back up traffic. As well, every intersection operates under the uncongested flow condition. The simulation results of the graph-based networked control of a multiple autonomous intersection showed that the overall traffic flow in the network achieves up to 20% higher than using traffic signal system.

On transient stability of multi-machine power systems through Takagi–Sugeno fuzzy-based sliding mode control approach

“…Schaab et al have considered robust control for voltage and transient stability of power grids focusing on wind power, and also Mazhari et al have addressed a frequency-domain approach for distributed harmonic analysis in multi-area interconnected power systems [3,4]. Darabian et al have presented a power control strategy, to improve power system stability in the presence of wind farms by designing predictive control and Shah et al have studied the performance improvement of intrusion detection with fusion of multiple sensors, while Wuthishuwong et al have focused on consensus-based local information coordination for the networked control of the autonomous intersection management [5][6][7]. In Yipeng et al's.…”

“…(5)- (7). However, the remaining equations are only constructive for short time constants, which are negligible in designing the stabilizer.…”

Adaptive Routing of Autonomous Vehicles Using Neighborhood Traffic

“…In simpler terms, during congestion scenarios, vehicles may have to follow the speed limit posted by the IM so that it will reach the intersection at a time that will not cause a jam in the outgoing street. The common goal between the work presented in [38] and ours is reduced wait times and reduced congestion in the network. While there are many ways to achieve this, driving at erratic speeds is certainly not a realistic approach.…”

“…In a trac system, the environment in which vehicles drive is a road network which is made up of link segments and nodes. The network of roads is usually modeled as an undirected graph for simplicity [38]. Nodes form the intersection.…”

“…al. present a similar routing algorithm based on I2I messages[38]. They divide the control present in the IM into two levels: network level and the control level.…”