Simulation Research On Traffic Flow Collaborative Optimization In Freeway Ramp Weaving Area

As an important part of the highway,the weaving area composed of inflow and outflow has long been of great concern to traffic managers and many scholars.Driven by the frequent lane change demand,the vehicles in the weaving area make the traffic flow operation state particularly chaotic,which is easy to cause traffic congestion.Therefore,the traffic flow optimization organization strategy for the weaving area has become the key to alleviating regional traffic congestion and improving the overall operation efficiency of the expressway.At present,the traffic flow optimization method in the weaving area is basically aimed at improving the driving efficiency,ignoring the inherent characteristics of the traffic flow in the area,such as high driving risk.Therefore,under the new opportunity of the development of Cooperative Vehicle Infrastructure System(CVIS),this paper considers the subjective preference and uncertainty of the driver’s decision-making behavior,and formulates the sub-regional dynamic lane change guidance control decision of the interwoven vehicle.Firstly,by analyzing the lane-changing behavior of vehicles in the weaving area,the characteristics of forced lane-changing behavior of weaving vehicles are determined.The exponential moving average filtering algorithm is used to denoise the NGSIM data set,extract the trajectory characteristics of the forced lane change of the incoming and outgoing vehicles,and analyze the traffic flow operation status of each lane in the weaving area.It is proposed that the forced lane change behavior of the weaving vehicle cluster aggravates the traffic congestion degree of the outer lane.In order to quantify the potential driving risk in the weaving area,the conflict types of various types of running vehicles are summarized and the evaluation quantitative indicators are formulated to provide theoretical support for the subsequent analysis of the driving risk level in the weaving area.Secondly,based on the decision-making psychology of the driver ’s bounded rationality,a cellular automaton dynamics model of the weaving area is established.In order to more accurately simulate the interweaving area with a large number of lane change behaviors,the driver ’s lane change perception stage is summarized.When establishing the model,combined with the improved prospect theory value function,the dynamic lane change distance perception of driver heterogeneity is introduced;in order to simulate the driver ’s forced lane change pressure,a lane change probability model with increasing weaving vehicles is constructed for the unique geometric structure of the ramp weaving area.In the establishment of the car-following model,the driver ’s estimated speed of the preceding vehicle is introduced.The model was corrected and the parameters were revised through the data sets.Then,a partition dynamic lane change guidance control strategy for traffic flow collaborative optimization is proposed.Starting from the conflict of vehicle operation in weaving area,a quantitative model of driving risk is established.Taking the minimum total travel time TTS of the vehicle running in the weaving area as the objective function,the ideal lane change guidance ratio of each partition area is obtained by the optimization algorithm.In order to consider the change of regional driving risk under the limited rationality of the driver,the mandatory lane change guidance measures and appropriate lane change guidance measures are proposed.The optimization algorithm is extended by the established feedback-penalty mechanism to reduce the risk of driving in the region.Finally,the cellular automata is used to build the simulation scene.By simulating and analyzing the heterogeneous behavior characteristics of drivers in the ramp weaving area from the spatial and temporal characteristics,lane change distribution and traffic flow characteristics of the weaving area,it shows that the dynamic model can effectively describe the bounded rational behavior of drivers in the weaving area.In order to further analyze the control effect of the optimization model,the non-control strategy and the dynamic lane change guidance control strategy are compared and analyzed from the aspects of regional driving efficiency and driving risk.Verify the effectiveness of the traffic flow optimization measures proposed in this paper.