Research On Cooperative Control Methods Of Traffic Flow At Urban Intersection In An Intelligent And Connected Environment

The transformation of intersection traffic management mode from traditional to intelligent and connected is an inevitable trend.Since the increase of Connected and Autonomous Vehicles(CAVs)penetration rate,the objects involved in intersection traffic and the essence of traffic control will be significant changed.The traffic flow at intersections will gradually transition from a mixed intelligent and connected environment where CAVs and Human-driven Vehicles(HDVs)coexist to an environment solely occupied by CAVs.Moreover,the control strategies of intersections traffic will be also progressively shifted from "passive responsive control" to "active cooperative control".On one hand,the changes in traffic participants at intersections lead to alterations in the intrinsic structure of traffic flow,posing new challenges to traditional research on microscopic vehicle behaviour analysis.On the other hand,conventional studies on intersection traffic management strategies have encountered insurmountable bottlenecks in efficiency.The input data largely depends on fixed detection equipment,providing traffic information that is often crude,limited,and lagging,making it difficult for implemented control strategies to respond to real-time changes in traffic conditions.The pressing issue in the development of intelligent intersections is how to analysis the operational mechanisms and control methods of traffic flow at intersections under such environments in order to adapt to the development of intelligent and connected trends.This study closely aligns with China’s urban traffic development needs and focuses on the key scientific problem of " research on cooperative control methods of traffic flow at intersection in an intelligent and connected environment".Considering the main stages of intelligence development in intersection traffic management,this study systematically unveils the differences in vehicle microscopic behaviours,traffic flow characteristics,and cooperative control mechanisms at typical intersections in each stage.Through research on vehicle microscopic behaviour analysis and simulation modelling,vehicle trajectory coordination with traffic signals,and sequential decision-making and trajectory optimization for vehicle passage,this study proposes cooperative control strategies for intersection traffic flow suitable for different stages of intelligent connected development.The research content encompasses: the design of a hierarchical framework for cooperative control of intersection traffic flow in an intelligent connected environment,analysis and modelling of microscopic behaviours for two types of vehicles,and cooperative control strategies for multiple typical intersection traffic flows.(1)This part aims to analyse the cooperative control problems for intersection traffic flow in an intelligent and connected environment and design a hierarchical structure framework.To reveal the mechanism of cooperative control for intersection traffic flow under intelligent and connected environments,this study conducts a detailed summary and analysis of related research on intersection control issues.It identifies the conflicts between intersection traffic management models and the development of intelligent and connected technologies,taking into account the unique characteristics of intelligent and connected environments,such as high complexity in traffic participants,large variations in traffic control scenarios,and broad dimensions of traffic evaluation indicators.From a systems theory perspective,a hierarchical structure framework is constructed,dividing the operation scenarios of intersection traffic flow into three development levels: early,middle,and late stages.Corresponding research methods for traffic flow cooperative control strategies are provided for each scenario,guiding subsequent studies on cooperative control methods for intersection traffic flow.(2)This part aims to analyse and model microscopic behaviours for two types of vehicles at intersections in an intelligent and connected environment.Addressing the unclear applicability of HDV microscopic behaviour models at intersection and the difficulty in leveraging CAV technology advantages,the paper constructs a generalized microscopic behaviour expression model for HDVs.By fitting model parameters based on the processing and screening of actual intersection data,a clear expression model for HDV microscopic behaviour is established.Subsequently,considering CAV’s perception advantage regarding the driving states of multiple preceding vehicles,the study examines the influence mechanism of factors,including vehicle type,multi-vehicle speeds,and headway distances on CAVs,and quantifies these traffic parameters based on the differences in communication reliability and vehicle motion approximation.This leads to the construction of a multi-perception microscopic behaviour control model for CAVs.Thereby it realizes the construction and validation of microscopic behaviour models for the two types of vehicles at intersections.Experimental results conducted in typical intersection scenarios demonstrate that the constructed models perform well in terms of intersection traffic efficiency and traffic flow stability across varying CAV penetration rates and lane-changing demand ratios,providing support for microscopic behaviour models in subsequent research on cooperative control of intersection traffic flow.(3)For cooperative control of signalized intersections in mixed intelligent and connected environments,the study proposes an Integrated-decentralized Bi-level Optimization Strategy.To address the issues with traditional signal control modes that fail to consider the characteristics of traffic flow in mixed intelligent and connected environments and the waste of spatial-temporal resources,the paper introduces a novel signal control method.Combining the safety and efficiency advantages of signal control and conflict elimination strategies,the intersection traffic control problem is decomposed into two subproblems: adaptive signal control and conflict detection activation.Green light signal optimization strategy and Red light signal optimization strategy are respectively constructed to realize integrated upper-layer and decentralized lower-layer control of intersection traffic signals.Furthermore,considering the complex interaction mechanisms and control requirements between CAVs and HDVs in mixed traffic flow environments,the study constructs models for leading vehicle trajectory control,following vehicle trajectory control,both for CAVs and HDVs.Ultimately,the strategy for intersection traffic flow is developed,achieving cooperative control of vehicle trajectories and signals in mixed intelligent and connected environments.The simulation experiments show that increasing intersection traffic demand and CAV penetration rates lead to increased activation of the WPCS strategy,reducing the waste of spatial-temporal resources at intersections.Additionally,algorithm complexity analysis reveals that IDBOS can meet the real-time demands of intersection traffic control.This study addresses the practical stage of intelligent intersection traffic control development(the early stage of intelligent and connected development).It is,in a traffic environment where CAVs and HDVs coexist,the presence of mixed traffic flow necessitates that intersections rely on traffic signals to coordinate the operation of the traffic flow.The adoption of the strategy allows for effectively enhancing traffic efficiency while ensuring the safe operation of traffic at intersections.(4)For cooperative control of signalized intersections in intelligent and connected environments,the study proposes an Analytical Trajectory Solving-Based Cooperative Control Strategy.To tackle the issue of simplified four-stage analysis of CAV trajectories resulting in reduced intersection traffic efficiency,the paper takes an approach from the perspective of optimal vehicle trajectory analysis and builds a multi-objective optimization model for intersection vehicle trajectories based on optimal theory.By categorizing the constraint violations of trajectory function’s control variables and state variables,the paper realizes the solution for the optimal trajectory of CAVs.Moreover,to leverage the controllability and communication capabilities of CAVs,the study uses the multi-perception microscopic behaviour control model to achieve more precise control over lane-changing and car-following behaviours.Based on the arrival information of vehicles at the intersection,a dynamic programming-based signal phase control optimization model optimizing traffic efficiency is constructed to realize the optimization of signal sequence and phase durations.Finally,by using an improved rolling optimization control strategy,the strategy is solved iteratively.Experimental results show that the proposed strategy not only effectively reduces vehicle delays at intersections but also minimizes energy consumption,pollutant emissions,while ensuring the stability of traffic flow.This research targets the transitional stage of intelligent transportation control development(mid-stage of intelligent and connected development),where traditional vehicles are being replaced by CAVs.However,since signal lights cannot be removed immediately,signal lights are still required to coordinate the passage order of CAVs,and the strategy achieves cooperative control between CAV trajectories and traffic signals.(5)For control of unsignalized intersections in intelligent and connected environments,the study proposes a Topology Diversity Analysis-based Multi-objective Optimization Strategy.Addressing the problem of vehicle conflict relationship analysis difference caused by the varying topological structures of intersection roads,this study considers the characteristics of optimal CAV trajectories and decomposes vehicle trajectory control into a two-stage optimal trajectory control problem.Using the first-in-first-out principle and vehicle conflict elimination rule under road grid division,a multi-objective optimization model is established with objectives including vehicle throughput efficiency,fuel economy,and driving comfort.Simultaneously,this study employs optimal theory and triggered constraint activation rules to analytically solve for the optimal vehicle trajectory.Experimental analysis shows that the proposed strategy effectively suppresses traffic fluctuations and stop-and-go phenomena,enabling vehicles to maintain high speeds in various topological structures of intersection areas.This research focuses on the ideal stage of intelligent transportation control development(late stage of intelligent and connected development),where CAV penetration reaches 100%,allowing for real-time sharing of all traffic information and vehicle trajectories.In this scenario,CAVs make cooperative decisions replacing the role of signal lights,and the strategy is used solely for the cooperative optimization of CAV passing order and trajectories.In conclusion,according to the development of China’s urban traffic and the direction of intelligent and connected development,this study considers the differences in traffic flow characteristics and cooperative control mechanisms at intersections across different stages of intelligent and connected development.It presents cooperative control strategies for intersection traffic flow suitable for various stages of intelligent and connected development,providing insights for studying typical intersection control problems and serving as a reference for future research on urban intersection traffic control strategies.