Cooperative On-ramp Merging Trajectory Planning And Motion Control For Connected And Automated Vehicles In Mixed Traffic

Cooperative on-ramp merging control for connected and automated vehicles(CAVs)is a potential approach to mitigate traffic conflicts and improve travel efficiency in the on-ramp merging zone,which is a typical traffic bottleneck.However,human-driven vehicles(HDVs)and CAVs will operate in the same environment for a long period of time in the foreseeable future.The stochastic and uncertain behaviors of human drivers bring huge challenge to cooperative on-ramp merging control of CAVs with regard to safety.This thesis aims to address the problem of cooperative on-ramp merging for CAVs in mixed traffic,and three research topics,including receding horizon merging trajectory planning and motion control strategies,event-driven merging mode switching and integration scheme,and scaled CAV test platform and experimental validation,are studied to explore the potential to improve vehicle driving safety,efficiency,and energy-saving effects,which further promotes the improvement of traffic efficiency and ecological performance.(1)Receding horizon cooperative merging trajectory planning for CAVs in mixed traffic.First,the cooperative on-ramp merging control scenario and multi-vehicle cooperation group are defined.Then,with the objectives of optimizing travel time and energy consumption,a control-constrained free-terminal optimal control problem(OCP)is formulated to render trajectories with flexible merging points.Further,the formulated OCP is embedded in a receding horizon optimization framework to tackle the uncertain behavior perturbation induced by HDVs,and the Pontryagin’s Minimum Principle(PMP)is employed to give optimal trajectory analytical solution.Simulation results in the typical on-ramp merging scenario show that the proposed strategy is robust and has the potential to optimize travel efficiency and energy consumption.(2)Safety-critical cooperative merging control strategy for CAVs.A nonlinear longitudinal vehicle dynamics model is firstly developed,and then a cooperative merging control problem considering constraints of dynamic limits and safe inter-vehicle distances is formulated.With the aim of minimizing travel time and energy consumption,a hierarchical cooperative merging control strategy is designed.The upper-level planner solves an unconstrained OCP with PMP to calculate an expected merging position,which provides the reference for the lower level.The lower-level controller converts the OCP with safety-critical constraints to a quadratic programming(QP)problem by exploiting Control Lyapunov Functions(CLFs)and Control Barrier Functions(CBFs),which transform the constraints and objectives into inequality constraints of the control input.The results of multi-vehicle simulations reveal that the proposed strategy can safely and efficiently realize merging control,and compared to the merging strategy with fixed merging point,this strategy can alleviate traffic shockwave and save 10% of traffic energy consumption.(3)Event-driven cooperative merging mode switching and integration control for CAVs.First,the control mode switching process for CAVs in on-ramp merging areas is analyzed,and an event-driven multi-mode control problem is formulated with mode switching condition design.Then,based on signal temporal logic(STL),an event-driven on-ramp merging control framework integrating vehicle model set and controller set is proposed,and the multi-mode control tasks of CAVs are converted to STL formulas.Moreover,the predicate functions of STL formulas can be mapped to CBF and CLF,and the CBF-CLF based multi-mode quadratic programmers are designed.Simulation results prove that the proposed event-driven cooperative control scheme is competent to make CAVs travel safely and efficiently in on-ramp merging areas under mixed traffic.(4)Scaled CAV test platform and experimental research.The hardware components,including a scaled smart city,scaled experimental vehicles,and motion capture system,are built.The cloud control software system,composed of communication,trajectory planning,and motion control modules,are designed.Single-vehicle experiments including accelerationcruise control and close-loop path tracking are conducted;what’s more,safe car-following and on-ramp merging control experiments are conducted to validate the effectiveness of the proposed event-driven receding horizon control strategy.