Research On Control Maneuver Of Intelligent Vehicle Lane-changing Behavior In Expressway Scene

Intelligent connected vehicle is the significant research field and development direction in China,and lane changing behavior takes the essential position in this technosphere.This paper studies the behavior control maneuver of automated lane change with the straight road section scene of expressway and the low speed obstacle vehicle ahead.The maneuver researches on the key technologies including behavioral decision-making,trajectory planning and control algorithms to realize the safe and comfortable lane changing of intelligent vehicles.Firstly,Based on the analysis of drivers’lane changing process and expressway driving environment,lane changing rules and lane changing decision-making model considering lane changing inducement and driving state of vehicles in front are established.Secondly,this paper chooses the curve interpolation method as the trajectory design theory.A five-order Bezier curve optimal lane-changing trajectory plan with the longitudinal positions x₁ and x₂ of the curve control points as undetermined coefficients is proposed.And the quintic Bezier and polynomial lane change curve are compared and analyzed.Compared with polynomial curve,quintic Bezier lane changing trajectory is smoother.Besides,quintic Bezier curve can be optimized by only obtaining the vehicle position relationship before planning.Therefore,the quintic Bezier trajectory is determined as the planning trajectory.Thirdly,a linear discrete state space is established based on the vehicle monorail dynamic model.Considering the constraints,the objective function is designed,and the optimal control sequence is obtained by transforming it into a quadratic programming problem.On this basis,the lateral control algorithm is established.As for the longitudinal control algorithm,the algorithm adopts a hierarchical structure whose upper layer obtains the demand acceleration by the MPC algorithm and the vehicle inverse model transforms the demand acceleration into the control quantity in the lower layer.In order to verify the tracking effect and stability of lateral control algorithm,the different speed conditions are used for simulation.The lateral tracking error of the control algorithm is within 0.129m,and the algorithm meets the control requirements.In addition,for verifying the performance of the longitudinal control algorithm,simulation experiments are carried out under the following acceleration and deceleration conditions.The maximum longitudinal deviation is2.01km/h,and the following effect is stable.Finally,a joint simulation scenario is built based on Car Sim/Simulink.The control maneuver of lane change is tested by setting the vehicle in front of the target lane under different working conditions.The experimental results show that the intelligent vehicle lane-changing behavior control maneuver can manipulate the vehicles to change into the target lane safely and reliably,and meet the functional requirements.