Research On Eco-driving Control At Signalized Intersections Under Connected And Automated Vehicles Environment

Eco-driving control is a traffic control method motivated by energy-saving emission reduction of vehicles.With the rapid development of communication and control technology,the feasibility and effectiveness of eco-driving control are gradually improved,it has good prospects in the prevention and control of road traffic air pollution.At present,different levels of eco-driving control on highway have been implemented in some parts of foreign countries to improve fuel efficiency and reduce emissions of vehicles.But for eco-driving control on signalized arterial,the correspongding theories are not feasible currently due to that most recent researches overemphasized the fully connected and automated vehicles environment and neglected the partially connected and automated vehicles environment consists of connected and automated vehicles(CAVs)and conventional human driven vehicles;meanwhile,the recent researches could easily lead to more serious traffic congestion at the signalized intersections since they excessively pursued energy-saving emission reduction of vehicles and ignored mobility;furthermore,the theoretical continuity of recent researches is poor,no eco-driving control theory system has been formed for vatious kinds of intersections.Based on the microscope car-following theory and the optimal control theory,this thesis conducted a study on eco-driving control at signalized intersections under mixed traffic environment consists of connected and automated vehicles(CAVs)and conventional human driven vehicles by following a manline of “conventional human driven vehicle motion prediction ? eco-driving control at isolated signalized intersection with single lane?sensitivity analysis?improvement oriented to continuous signalized intersections with single lane?improvement oriented to signalized intersections with multiple lanes”,its control object is straight-going vehicle.Based on microscope car-following theory,this thesis adopted the concept of “Virtual Lead Vehicle” to improve various microscope car-following models of continuous traffic flow to orient interrupted traffic flow at signalized intersections,and utilized measured vehicle motion data to calibrate the parameters of each improved model.The improved model which has highest accuracy was chosen as the motion prediction algorithm of conventional human driven vehicle under connected and automated environment consists of connected and automated vehicles and conventional human driven vehicles,and provided basic road traffic perception support for following eco-driving control research.An eco-driving control system was designed for isolated signalized intersection with single lane.This system optimized the entire traffic flow by optimizing speed profiles of CAVs,so as to achieve the control objective of improving fuel efficiency and reducing emissions of vehicles while guaranteeing mobility at the intersection.The velocity optimization of CAVs was solved using Pontryagin's Minimum Principle.The solving process was accelerated by bi-directional iterative algorithm to enhance the timeliness of control system.An extensive test on the eco-driving control system at isolated signalized intersection with single lane was conducted to evaluate the control effect of this system in various networks under the influence of various factors.Sensitivity analysis method was adopted to conclude the application conditions of this system implemented at isolated signalized intersection with single lane,which can provide guidance for system implementation in the real world;the analysis also explored the feasibility of this system implemented at continuous signalized intersections with single lane,and found out its adverse control effect and the traffic mechanism behind this effect,which provided direction guidance and theory support for continuous signalized intersections with single lane oriented improvements of this system.The eco-driving control system at isolated signalized intersection with single lane was improved so as to orient continuous signalized intersections with single lane.The adverse control effect of eco-driving control system at isolated signalized intersection with single lane implemented at continuous signalized intersections with single lane was suppressed by modifying the optimal speed trajectories of CAVs.A vehicle benefits balance optimization model with the goal of maximizing system ecological benefits was constructed to reduce the loss of fuel consumption benefits caused by adverse effect suppression.The velocity optimization of CAVs was improved using offline optimization method to further enhance the timeliness of control system.The eco-driving control system at continuous signalized intersections with single lane was improved by introducing vehicle cooperative lane-change control so as to orient signalized intersections with multiple lanes.The velocity optimization model for vehicle cooperative lane-change was built based on the optimal safety clearance adjustment logic,which can improve fuel efficiency while ensuring priority of vehicle lane-change motion.A velocity optimization for vehicle motion after lane-change was constructed to improve fuel efficiency after vehicle's lane-change motion while maintaining mobility at the intersections.An eco-driving control was formed for the whole process of before-during-after lane-change motion.At last,all these three eco-driving control systems were applied on some urban roads of Harbin based on the Pre Scan/Simulink simulated driving platform,each system showed good timeliness,effectiveness and advancement.In this thesis,the research of eco-driving control atsignalized intersection was conducted under connected and automated vehicles environment consists of connected and automated vehicles and conventional human driven vehicles,an eco-driving control theory system was completed from isolated intersection to continuous intersections,and from single lane to multiple lanes.This thesis can provide scientific basis and theoretical support for energy-saving emission reduction of road traffic.