Research On The Optimal Deployment Of Dedicated Connected Autonomous Vehicle Lanes

In recent years,connected autonomous vehicles(CAVs)have received unprecedented attention from the government,industry,and academia.The development of CAV has a positive impact on the transportation system in many aspects,including improving road safety,increasing road capacity,improving transportation operation efficiency,and lower energy consumption.Promoting the development of CAV can provide means for solving traffic problems such as traffic accidents,environmental pollution,and urban congestion.Although CAV is developing rapidly,it will take a long time to be widely adopted.The mixed traffic flow composed of human-driven vehicles(HV)and CAV will inevitably exist.Mixed traffic flow with CAV will change the composition of traditional traffic flow and brought new traffic characteristics.Affected by the development of CAV technology,there are many uncertainties of the mixed traffic flow with CAV.To give full play to the advantages of CAV,deploying dedicated CAV lanes has become an important technology trend and management choice.The provision of dedicated CAV lanes will help CAV save travel time,further increase the market penetration rate of CAV,and reduce system delays.At present,deploying dedicated CAV lanes is an effective way to improve traffic efficiency and expand the market penetration of CAV.On the other hand,deploying dedicated CAV lanes should also consider the impact of CAV on the traffic environment and equity.This paper focuses on the optimal deployment problem of dedicated CAV lane,and conducts some researches from the following aspects:(1)Basic formulations on the deployment of CAV dedicated lanes.First,a method for calculating the capacity of the lane with mixed traffic flow is proposed considering the influence of the penetration rate of CAV;Secondly,a generalized travel cost function including travel time,fuel,and other factors is constructed to split the travel modes between CAV and HV.And the nonlinear complementary conditions of the mixed traffic flow with CAV multi-user equilibrium are derived.Finally,based on the investigation of road network design problems,the research framework of the bi-level planning model for the deployment of dedicated CAV lanes is constructed,which provides a theoretical foundation for subsequent research.(2)Joint deployment optimization of CAV lanes and bus lanes.Given the low penetration rate in the early stage of CAV development,the low utilization rate of bus lanes,and road resources,the joint deployment optimization of CAV lanes and bus lanes is proposed.For the joint optimization problem,a bi-level programming model is established with the goal of optimal system travel time.A solution framework combining an artificial bee colony algorithm and MSA is proposed to solve the model.Finally,based on two examples of the Nguyen-Dupuis network and Sioux Falls network,numerical analysis and parameter sensitivity analysis of three scenarios are carried out to verify the feasibility of the model and algorithm.Compared with the network without dedicated lanes and only bus lanes,the joint layout method can effectively reduce the travel time of the system.(3)Research on the deployment of dedicated CAV lanes for sustainable development.Sustainable transportation development requires a balance between the transportation environment and transportation efficiency.To analyze the impact of the establishment of CAV dedicated lanes on sustainable transportation,a multi-objective optimization problem considering the system travel cost,dedicated CAV lane construction cost,and system emission cost is proposed.For this problem,a multi-objective bi-level model is constructed.The upperlevel planning model seeks the optimal solution for multiple goals of sustainable development and traffic efficiency,and the lower-level model is a multi-user equilibrium problem.For this model,a solution framework that combines NSGA-II,diagonalization algorithm,and FrankWolfe algorithm is proposed to solve the model.Finally,based on the Nguyen-Dupuis network and Sioux Falls network,a case analysis and sensitivity analysis of model parameters is carried out to verify the feasibility of the model and algorithm.Based on the analysis results,deploying dedicated CAV lanes can effectively reduce system emissions and promote the development of sustainable transportation.(4)Research on dynamic planning of dedicated CAV lanes considering equity.Considering the uncertainty in the development of CAV,the dynamic planning of dedicated CAV lanes is proposed.First,to better capture the changes in CAV demand,a generalized Bass model that considers CAV prices and economic factors is established to reflect the changes in CAV demand with the planning year.Secondly,to study the influence of dedicated CAV lanes on equity,an equity evaluation method is proposed.A multi-objective two-stage dynamic programming model considering equity and system travel cost is established,and the multi-objective optimization problem is transformed into a single-objective optimization problem with two optimization orders by using the dictionary sorting method.For this model,a solution framework combining artificial bee colony algorithm,variable neighborhood search algorithm,and i TAPAS algorithm is proposed.Finally,based on the Nguyen-Dupuis network and Sioux Falls network,a case analysis and sensitivity analysis is carried out to verify the feasibility of the model and algorithm.Based on the analysis results,dynamic planning considering equity is beneficial to reduce system travel time and improve equity.