Research On The Shortest Path Algorithm Under Occasional Congestion Based On Spatial Correlation Of Traffic Flow

Travel time is the main factor affecting the travel satisfaction of residents,while traffic congestion increases the travel time of cars and reduces the travel efficiency,thus affecting the travel satisfaction of residents.Traffic congestion can be classified into frequent congestion and occasional congestion.The occasional congestion is characterized by randomness and non-periodicity,its direction and extent of propagation,degree of impact are more difficult to predict.When occasional congestion occurs,the commonly used shortest path algorithm does not take into account the spread of congestion which causes dynamic changes in the traffic state of the road network.When recommending the path,maybe it is impossible to avoid the potential congestion.In response to this problem,this paper studies the spatial correlation between traffic flow on the links,and portrays the propagation process of occasional congestion in the urban road network.According to the change of the traffic status of the road network with time,the route with the shortest travel time is searched to guide the vehicle to reduce travel delays.The main research work completed by the thesis is as follows:(1)Research on Data-driven spatial correlation of traffic flow on urban road networkTaking the processed GPS trajectory as the basic data,the Apriori algorithm is used to extract the segments and paths that frequently appear in the trajectory,and the association rules are generated.The confidence of the correlation rule is taken as the coefficient to represent the spatial correlation between the traffic flow,in order to describe the strength of the interaction between the traffic flow.(2)Analysis about traffic flow state of links mainly affected by occasional congestionThe traffic flow theory is used to analyze the traffic flow state of the link where the event occurs.The volume and speed changes of the original link under different lane failure scenarios are obtained by VISSIM simulation.According to the spatial correlation between traffic flow,we can quantify the traffic flow changes in the upstream and downstream links to analyze the changes of traffic flow status of the related links.(3)Travel time prediction under occasional congestion considering spatial correlation of traffic flowConsidering the spatial correlation and the dynamic process of traffic flow ineach section of the road network,the time is discretized by the sudden change of traffic state of the road section,and then the volume of the road section is dynamically loaded.Based on the traditional traffic impedance model,the transit time of the road segment is transformed into a function related to Departure time,volume,and capacity,and a travel time prediction model under occasional congestion is established.(4)The shortest path algorithm considering the process of congestion propagationEstablish the topology network of urban roads.Use the travel time of the road section that changes with time as the dynamic road resistance.Based on the improved Dijkstra algorithm,with the shortest travel time as the goal,calculate the shortest path between the original and destination of travel after the congestion occurs.Compare this path with the shortest path obtained by the conventional algorithm without considering the congestion propagation process,which verifies the important impact of the congestion propagation on travel time.