Design And Implementation Of Traffic Situation Evolution System For Digital Twin

With the rapid development of urbanization,traffic congestion has become one of the urgent problems to be solved in big cities.It is an important way to solve the problem of traffic congestion by planning a suitable path for vehicles to avoid time-space aggregation under the condition of travel time constraints.For this reason,the industry puts forward the traffic digital twin,which aggregates the traffic status by recognizing and collecting the vehicle location to support vehicle path planning.However,due to the coupling between the vehicle route planning decision and the traffic situation,the current route planning results of a large number of vehicles finally emerge the traffic state.In order to optimize,all potential paths of all vehicles need to be enumerated.With the expansion of the transportation network and the increase of vehicles participating in the planning,the search space will become very large and difficult to calculate.In order to solve this problem,this paper proposes a traffic situation deduction algorithm based on digital twinning,which models large-scale paths into a Markov process and evaluates the traffic state through deep learning.In order to solve the problem of vehicle scale,firstly,the road network is divided into multiple regions,and the processing scale of each planner is reduced through hierarchical regional planner and global planner.The regional planner is responsible for coordinating the path planning of vehicles in the region and generating a variety of different regional guidance strategies.The global optimizer is used to evaluate the combination of various strategies to balance regional equilibrium and system optimization.In the planning process,the algorithm uses Monte Carlo tree search to dynamically access and simulate the future state,so as to reduce the search space,and reduce the dimension by pruning to achieve more efficient search.Simulation analysis shows that the algorithm proposed in this paper can plan a better path for all vehicles in the road network in a limited time,and reduce the congestion time of the road network.Based on the traffic situation deduction algorithm,this paper designs and implements a digital twin-oriented traffic situation deduction system,which is used to support the traffic situation evolution analysis in the traffic digital twin system.In this paper,the function of traffic situation deduction is analyzed,and the system is divided into data access layer,situation deduction layer and user service layer.The data access layer is responsible for importing data from the real world and is used to create digital road networks and digital traffic flows;the situation deduction layer is used to execute the traffic situation deduction algorithm to plan the path for the vehicles in the road network;the user service layer is used to visualize the situation deduction results and verify the deduction results.Finally,the paper designs detailed test cases for the functional points of the system and carres out functional tests to verify the correctness and availability of the system.