Study On Regional Coordination And Control Of Roundabouts Based On Macroscopic Traffic Flow Model

The roundabout appeared in the 19 th century,it attracted many domestic and foreign scholars and engineering and technical personnel to pay close attention due to its high traffic efficiency and low traffic accident rate.At medium and low saturation,vehicles in roundabouts follow traffic rules of self-organization,and the capacity of the roundabouts far exceeds that of ordinary intersections;however,at high saturation levels,traffic jams occur in roundabouts,and roundabouts become the bottleneck of urban road network.The macroscopic traffic flow model of roundabouts mainly studies the evolution of traffic flow in roundabouts,by deepening the essential characteristics of traffic flow in roundabouts,and analyzes the mechanism of traffic congestion,can also provides the theoretical basis for both the microscopic simulation of roundabouts and the establishment of control strategies.The optimizing method of regional control for roundabouts considers the operational characteristics of traffic flow,and thinks the region as a whole,it effectively controls and optimizes the design of the intersection within the region,making the regional road network play its role.Therefore,establishing the macroscopic traffic flow model,the travel time prediction model and raising the control strategy of the roundabout is of great significance to the management?planning and the traffic flow theory of the roundabouts.This paper first toke the cross section of a weaving section as the reference point,analyzed the relationship between the release of the previous weaving section,the entry of the entrance lane and the flow through the weaving section,and established the maximum release model of the entrance lane under the maximum number of vehicles passing through the weaving area.Considering the influences of in-and-out roundabout ratio of the weaving section on vehicle operation,the average speed model of the weaving section was provided.Secondly,combining the survey data,this paper calibrated the boundary of roundabout among entering area,circle area and exiting area using the cumulative probability of the lane-changing initial location;Established a relation model between the probability of lane choosing and volume with field data;Considering roundabout geometry condition,vehicle queuing and lane changing,a travel time prediction model was established obtaining entering time,circle time and exiting time.Finally,this paper analyzed the traffic conditions in the lanes and weaving areas,and selected and established the state indicators of road sections and weaving areas and the associated index models of weaving sections;Describing the traffic conditions existed in the area and proposing corresponding control strategies;Combining the traffic flow model of the weaving area and the roundabout travel time prediction model,a heuristic dynamic programming(HDP)model based on BP neural network was further designed to solve the regional coordination model.Based on the regional coordinated control model,the JAVA language was used to build the algorithm platform,and the road network parameters were set and the data was collected.The effectiveness of the coordinated control model was proved by comparing the model calculation and the VISSIM simulation to get the number of outbound vehicles.Then the control parameter solving process was given by establishing the HDP model,and the regional network signal was calculated and dynamically optimized based on the BP neural network.Finally,combined with the VISSIM simulation software,four performance evaluation indexes under the fixed road network and the control method proposed in this paper was compared,by which verified that this control strategy can effectively prevent the deadlock phenomenon in the roundabout and realize the regional coordinated control of the roundabout.