Research On Bottleneck Optimization Strategy In Confluence Section Of Expressway Based On Cooperative Control

As the main skeleton of the urban road network,expressway provides efficient and convenient transportation services for long-distance urban travel,and plays an important position in urban roads.However,the confluence bottleneck section often becomes congested and accident-prone areas due to the conflict between the mainline and the onramp traffic streams.With the continuous development and maturity of transportation information technology,traffic control has been further optimized.By combining multiple traffic control strategies,it has greater advantages in alleviating confluence bottleneck and improving traffic efficiency.Among them,ramp control and variable Speed Limit(VSL)control are the most widely used.Therefore,taking the bottleneck section of urban expressway confluence as the research object,a multi-lane traffic flow model is established,and based on the Model Predictive Control(MPC)framework,a cooperative control strategy for ramp control and Variable Speed Limit control is proposed.Firstly,traffic data are obtained by video recording and manual recording,selects headway and speed for error analysis.For confluence bottleneck section,the mainline traffic flow characteristics are analyzed based on three traffic flow parameter characteristics,lane changing characteristics,and headway characteristics.For confluence traffic flow characteristics,the on-ramp traffic flow characteristics and the confluence-gap characteristics are analyzed.The analysis shows that during peak hours,the speed and flow of each lane on the mainline decrease from inside out,and vehicles keep moving and stopping,which can easily cause congestion.The outer lane is disturbed by uneven traffic flow on-ramp,and tend to change lanes towards the inner side of the upstream confluence area,and the distribution of headway is discrete.Moreover,the confluence gap between the ramp traffic flow and the acceleration lane driving distance shows a linear relationship based on the critical gap analysis.Secondly,based on the analysis of traffic flow characteristics,the classic Cell Transmission Model(CTM)is modified.The basic cells,diffluence cells,and confluence cells of the classic CTM are improved from three aspects,cell length division,traffic flow fundamental diagram,and on-ramp cell update.Based on the improved CTM,the lanebased CTM model is constructed by introducing the probability of changing lanes under the influence of density difference,speed difference,and density difference with blocking density.Then,the capacity model of on-ramp is set up under the gap acceptance theory of variable critical gaps,and the ramp control strategy is formed with the ALINEA algorithm that considers ramp queues.The on-ramp control and VSL control are embedded into the lane-based CTM model to establish a control system and optimization model for ramp and VSL control strategy.Based on the MPC control framework,a cooperative control strategy is designed,and the critical density and travel speed are used as the starting conditions for the control strategy.Finally,case simulations are carried out on the VISSIM platform,simulation parameters and model control parameters are calibrated.And MATLAB is used for Secondary development to establish traffic scenarios under no control,ramp control strategy,VSL control strategy,and cooperative control strategy.Comparisons and analyses are conducted on the local control effect and the overall traffic efficiency of the road network.The results show that compared to the uncontrolled state,the cooperative control strategy optimizes the total travel time by 9.24%,increases the total turnover mileage by2.59%,and improves the total delay time by 17.52%.Compared with other control strategy,the cooperative control strategy can better improve the traffic efficiency of confluence bottleneck section,make the traffic flow smoother,turnover volumes increased,and total travel time and delay time reduced.