Intelligent Vehicle Trajectory Planning And Tracking Control With Multi-performance Object Collaborative Optimization

The trajectory planning and tracking control of the key technologies of intelligent vehicles determine the safety,economy and comfort of the vehicle.Aiming at the multiobjective optimization problem of planning and control.It can effectively improve the vehicle’s intelligence level that multi-performance target optimization and high realtime algorithm research of lane change trajectory planning and trajectory tracking control of intelligent vehicle on structured road.Combined with part of the research content of the National Key R & D Program(No.2016YFB0100905)and Chongqing Natural Science Foundation(No.cstc2018 jcyj AX0422).Considering the lane change trajectory planning and trajectory tracking control in the complex scene of multi-vehicle motion,with the aim of improving the vehicle’s safety,comfort,economy and real-time performance of control algorithms,a structured road lane change trajectory planning and explicit model predictive control algorithms for trajectory tracking was proposed.(1)The trajectory planning of structured road changes.In order to realize the lane change trajectory planning of intelligent vehicles under the complicated traffic environment on structured roads and improve the safety,smoothness and economy of lane change trajectories,a lane change trajectory planning algorithm based on multiobjective optimization was proposed.The trajectory planning of curved roads was transformed into the trajectory planning of straight roads by the frenet coordinate system.Trajectory planning was transformed into two low-dimensional planning problems: trajectory shape planning and speed planning.The shape of the lane change trajectory of the vehicle was planned by an improved artificial potential field algorithm.An important parameter was predicted by the GA-BP neural network based on the NGSIM data set to improve the anthropomorphic degree of lane change.In longitudinal speed planning,the mathematical representation of the surrounding environment was established based on the displacement time S-T diagram.The improved RRT algorithm was used to achieve the optimal speed planning under known track parameters,and the effectiveness of the trajectory planning algorithm was verified by simulation.(2)Trajectory tracking control based on multi-objective optimization.In order to realize trajectory tracking control and ensure multi-performance optimization in the tracking process,a trajectory tracking control method based on explicit model predictive control theory was designed.The tracking accuracy,driving stability and comfort were quantitatively analyzed.The trajectory tracking control problem was divided into two sub-problems: longitudinal flowing control(LOF)and lateral flowing control(LAF).Based on the explicit model predictive control theory(EMPC),the online iterative optimization solution problem in the traditional model predictive control system was converted into an equivalent polyhedron segmented affine system equivalent to it,and the optimal amount of control was gain by the explicit control law on the parameter partition.EMPC-LOF and EMPC-LAF controllers were designed for longitudinal and lateral tracking control respectively,and the effectiveness of the control method was verified by simulation.(3)Co-simulation of lane change trajectory planning and trajectory tracking.A trajectory planning and trajectory tracking control co-simulation verification method based on MATLAB / Simulink and Carsim simulation software was provided,and the integrated performance of the trajectory planning and tracking control algorithm under complex lane change scenarios was verified by the co-simulation platform.