Research And Simulation Of Dynamic Route Recommendation Method Based On Traffic Situation Awareness

With the increase of car ownerships year by year,urban traffic congestion occurs frequently.One of the important reasons for the deterioration of traffic environment is that drivers can not obtain real-time,accurate,all-round and efficient traffic service information.It is urgent to integrate communication,sensing control and computer technology into transportation,establish an intelligent transportation system,and realize dynamic route guidance.The optimal path planning method adopted by traditional intelligent transportation system does not fully consider the changing situation of urban traffic,which makes the planned path unable to cope with the changes of traffic environment,thus the user experience is poor.In order to solve this problem,this paper proposes a dynamic route recommendation method based on traffic situation recognition.This method uses massive taxi GPS trajectory data for cognitive learning,and dynamically recommends travel routes to users.The method consists of three parts:1)Considering the temporal and spatial factors affecting the urban traffic condition,a short-term traffic situation prediction model based on residual network structure is proposed.The short-term traffic situation change is predicted by using the cyclical trend of traffic condition.2)Considering the decision-making experience of taxi drivers,a value iteration grid network oriented to travel planning is proposed,which learns taxi drivers’experience knowledge of road-finding from taxi trajectory.Based on the value iteration grid network,a driving decision model with situation awareness is designed.According to the value iteration grid network and the short-term traffic situation prediction,the coarse-grained grid path with situation awareness is planned.3)On the basis of coarse-grained grid path,the equivalent static road network model is obtained by time expansion,and the travel time of real-time grid path planned by driving decision model is used as the cost estimation in heuristic search.A*dynamic path planning algorithm based on traffic situation-aware is proposed.Finally,simulation experiments are carried out on the proposed urban short-term traffic situation prediction model,situation-aware driving decision model and A*dynamic path planning algorithm based on traffic situation-aware.The experimental results show that the proposed method is reasonable and effective.