Research On Collaborative Confluence Method Of Expressway Ramp Under Intelligent Network Environment

As a typical traffic bottleneck area,the ramp confluence area not only has the problem of low traffic efficiency,but also has the potential traffic safety risk that the main road vehicles near the ramp side may have more aggressive driving behavior when there are vehicles for confluence operation.With the rapid development of related technologies such as intelligent connected and vehicle infrastructure coordination,through the information interaction between intelligent connected vehicles and infrastructure,drivers can timely obtain the traffic conditions in the surrounding areas.Therefore,with the help of intelligent connected technology,combined with collaborative confluence control strategy,the traffic environment of ramp confluence area can be improved by sending control information.However,under the premise of improving the traffic environment of the ramp confluence area,how to adopt the above strategies to be more accepted by the driver and how to reduce the amount of interactive information between the vehicle and the infrastructure are what we need to consider.Therefore,this thesis designs the corresponding collaborative confluence method from the above perspective.The main research work is as follows:(1)The analysis of vehicle operation characteristics in expressway ramp confluence area was carried out.In order to study the driving behavior of the driver in the ramp confluence area and analyze the influence of the confluence behavior of the vehicle on the traffic state of the main road.Based on the natural driving data set,the confluence point and confluence speed distribution of the vehicle,the relationship between the confluence time distance and the speed difference of the front vehicle are analyzed.Then,based on different scenarios,the rapid deceleration and speed distribution of the main road vehicle and its interaction with the front confluence vehicle are described.Finally,the genetic algorithm is used to fit the density-speed curve and analyze the traffic flow parameters of the main road.The above analysis verifies that the confluence behavior of the vehicle will have a negative impact on the operation of the main road vehicle.(2)A vehicle cooperative confluence control system for intelligent connected scenarios is designed.In order to improve the traffic operation efficiency of the ramp confluence area and reduce the traffic safety risk of the ramp confluence area,a cooperative confluence control system and overall strategy based on the intelligent connected scenario are proposed.Firstly,the development status of vehicle-road cooperative technology is introduced.Then,the system framework,specific design method of cooperative confluence control system and information transmission process are expounded.Finally,from the perspective of reducing the number of vehicle coordination and reducing the distance of vehicle coordination,the overall strategy of distributed-heuristic ramp cooperative confluence control is proposed in a relatively real road network environment.The strategy mainly includes two parts: the main road vehicle pre-lane change strategy and the speed spacing adjustment strategy of cooperative vehicles.(3)A vehicle cooperative confluence control method for intelligent connected scenarios is established.In this thesis,the overall control strategy of cooperative confluence is introduced in detail.Firstly,the MOBIL lane change revenue model is used to select the vehicle with the largest lane change revenue in the main road near the ramp side for lane change.Secondly,the real-time prediction model and the data-driven carfollowing model based on Bi GRU are designed,so that in the subsequent simulation process,the lane change revenue can be accurately calculated while the simulation vehicle is closer to the actual operating state.Then the cooperative vehicle and confluence spacing are selected.Finally,the selection rules and speed spacing adjustment rules of cooperative vehicles are established based on different confluence scenarios.On this basis,the car-following distance of merging vehicles is determined.(4)The simulation experiment of vehicle cooperative confluence for intelligent connected scenario is carried out.In order to verify the superiority of the real-time prediction model and the Bi GRU-based data-driven car-following model in speed prediction and the effect of the cooperative confluence control strategy in optimizing the travel time of the vehicle and the transportation performance of the ramp confluence area.Firstly,the training process of real-time prediction model and data-driven car-following model based on Bi GRU is introduced.Then,the above model is compared with other machine learning models and model-driven car-following models.The results show that the performance of the model proposed in this paper is better than other comparison models.Then,the above model is integrated into the collaborative merging strategy,and the traffic scenarios under different traffic flow and strategy information compliance rates are simulated by Vissim-python joint simulation.Finally,the simulation results are analyzed.The simulation results show that the strategy can optimize the travel time of the vehicle and the transportation performance of the ramp confluence area in the above scenarios.