Intersection Coordination For Autonomous Vehicles In Intelligent Transportation Systems

With the rapid development of industrialization and urbanization,more and more vehicles will flood into the road traffic system.This will bring a series of problems such as traffic congestion,frequent accidents and environmental pollution.With the development of the Internet of Vehicles in recent years,2020 is recognized as the first year of driverless technology.Unmanned driving technology is given high hopes to solve a series of problems that may be faced in the future intelligent transportation system,with the injection of a large amount of funds and research.Intersections are the confluence of different vehicle trajectories on multiple roads.Although they occupy a small proportion of the road,they usually become the bottleneck of the transportation system.At the same time,due to its unique characteristics of dense vehicles and complex environment,it also brings huge challenges to unmanned driving technology.Therefore,intersection coordination of unmanned vehicles has also become a hot spot in the field of unmanned driving in recent years.When designing the intersection dispatching system,not only must the safety of dispatching be ensured,but also the efficiency of the dispatching system must be improved as much as possible to give full play to the advantages of unmanned vehicles.However,the current research on unmanned vehicle intersection scheduling is mainly focused on the most basic single-vehicle intersection scene.There is still a lot of room for improvement in the throughput of the scheduling model.And there are relatively few researches on dynamic unmanned vehicle intersection scheduling.This paper proposes a variety of safe and efficient unmanned vehicle intersection coordination schemes for complex intersection areas,starting from the perspective of vehicle control.First,by pre-designing the trajectory of unmanned vehicles passing through the intersection,a fixed trajectory intersection coordination model is established based on five-collision-set.Based on the intersection model,static single-lane and multi-lane intersection coordination strategies are respectively proposed.Then,based on the time-varying characteristics of the vehicle’s dynamic arrival at the intersection,two dynamic intersection scheduling strategies based on five collision sets in time domain and spatial domain are respectively proposed.In addition,in order to further improve the throughput of intersections and maximize the use of space resources in the intersection,this paper also proposes a novel no-fixed trajectory unmanned vehicle intersection scheduling scheme based on a two-degree-of-freedom bicycle model.The simulation verification is carried out for the above-mentioned coordination strategy.The simulation results show that the intersection scheduling strategy proposed in this paper under the premise of ensuring safety in static scenarios has relatively large advantages in scheduling throughput,algorithm complexity and model scalability.In dynamic scenarios,the proposed coordination strategy can achieve efficient vehicle coordination under the premise of ensuring safety and system stability,and it is verified that the proposed spatial domain algorithm can be very close to the theoretical optimal solution of intersection coordination.The non-fixed trajectory intersection coordination strategy is not only suitable for various complex multi-lane intersection scenarios,but also can shorten the scheduling time by about 8.8% compared with the fixed trajectory scheduling scheme,which further improves the throughput of the coordination system.