|
Based on Vehicle-to-Vehicle, Vehicle-to-Infrastructure and Infrastructure-to-Vehicle communications, we propose a VLC system for managing vehicles crossing a light controlled intersection in a safe manner. Connected vehicles and infrastructure interact by broadcasting information using headlights, streetlights, and traffic signals. Transmitters emit light signals encoded, modulated and converted from data. Optical sensors with light filtering properties are used as receivers and decoders. A joint transmission allows mobile optical receivers to collect data, calculate their location for positioning, and read the transmitted data at the same time. A communication scenario is stablished. Parallel to this, an intersection manager coordinates traffic flow and interacts with vehicles through embedded Driver Agents. To command the passage of vehicles crossing the intersection safely queue/request/response mechanisms and temporal/space relative pose concepts are used. A dynamic phasing diagram and a matrix of states based on the total accumulated time are presented to illustrate the concept. On a Simulation of Urban MObility simulator (SUMO), deep reinforcement learning was used to control the cycle of traffic lights. Data shows that the adaptive traffic control system in the V2X environment can collect detailed data, including vehicle position, speed, queue length, and stopping time. Dynamic control of traffic flows at intersections is demonstrated using sequence state durations, phase diagrams, and average speed measurements. For the same traffic flow, static and dynamic cycle lengths were compared. According to the results, the dynamic system finishes the cycle first by adjusting the durations of the cycles as necessary. The better temporal management of phases results in better traffic flow and a higher average speed.
|
