Coordinated Optimal Control Of Internal Boundary And Ramp Inflows For Lane-free Traffic Of Connected And Automated Vehicles On Freeways

As an important part of traffic regulations in human driving environment,lane division is undoubtedly beneficial to traffic safety,but they also bring negative impacts,such as reducing the utilization rate of road lateral spatial resources.For driving safety can be guaranteed in high-level connected and automated driving environment,it is advisable to consider the use of lane-free traffic mode of connected and automated vehicles(CAVs),which can not only eliminate the negative impact of lane division,but also fully utilizes the technical advantages of CAVs,and promote a new form of traffic control,internal boundary control.In traditional human driving mode,the total spatial resources of roads are generally divided equally or evenly between the bi-directional traffic flows,which cannot flexibly cope with the mismatch between road resources and traffic demand.In the lane-free traffic mode of CAVs,the hard boundary between the bi-directional traffic flows can be ‘softened’.By high-precision regulation of the internal boundary,all bi-directional road resources can be scheduled to achieve the best match between traffic demand and road capacity in each direction and optimize traffic efficiency.This thesis focuses on lane-free traffic of CAVs on freeways,studies the internal boundary control mode and strategy,explores the advantages and limitations of internal boundary control,and its association with traditional highway control methods,and studies the coordinated control of internal boundary and ramp inflows.The main research contents and results include:(1)Construction of the lane-free traffic flow model of CAVs: The existing macroscopic traffic flow models are all aimed at the traditional lane-based mode,but cannot characterize the dynamic characteristics of the internal boundary under the lanefree traffic mode of CAVs.Firstly,this thesis constructs a dynamic model of the internal boundary of freeways,which characterizes the dynamic characteristics of the key parameters such as capacity,critical density,and congestion density based on the location coefficient of the internal boundary.Secondly,by combining the dynamic internal boundary model with the first-order macroscopic traffic flow model CTM with the ability to reproduce the capacity drop,an internal boundary adjustable macroscopic traffic flow model which is suitable for the control research of lane-free traffic of CAVs on freeways is established.The simulation results show that the model can well reproduce the capacity drop,dynamically set the internal boundary,and make the traffic flow propagate according to the corresponding internal boundary.(2)Design of the optimal internal boundary control model: Based on the proposed macroscopic traffic flow model,with the goal of optimizing the overall efficiency of the road network,and taking into account factors such as control stability,user rights balance,and so on,an optimal internal boundary control model for lane-free traffic of CAVs on freeways is constructed.Considering computational efficiency,the nonlinear constraints are linearized,and the holding-back phenomenon caused by linearization is eliminated by strengthening the objective function.The simulation study shows that the internal boundary control can well deal with the congestion caused by the imbalance of demand for the bi-directional traffic flow.(3)Design of the coordinated optimal control model for internal boundary and onramp in-flows: When the total bi-directional traffic demand exceeds the total capacity of the road,the congestion cannot be eliminated by internal boundary control alone,so it is necessary to introduce additional control measures.This thesis explores the feasibility of coordinated control of internal boundary and on-ramp in-flows and designs the optimal coordinated optimal control model.The simulation results show that,for a given freeway scenario without loss of generality,the internal boundary control can make full use of road spatial resources to cope with the traffic demand in each direction,and make the total travel time(TTS)decreased by 41.8%,but cannot avoid the capacity drop in the bottleneck during peak periods.Ramp metering can ensure that there is no capacity drop in the on-ramp bottleneck area,resulting in a decrease of 16.1% of TTS,but cannot deal with the mismatch between capacity and traffic demand in each direction.The coordinated control of internal boundary and onramp in-flows has both advantages,which reduces TTS by 43.5%.What’s more,it can also realize the sharing of the opposite ramp queuing space in the case of limited ramp space.