The Research Of Velocity Planning And Lateral-longitudinal Control Method For Intelligent Vehicle

With the development of Chinese automobile industry,a large number of vehicles have increased the problem of exhaust pollution and traffic congestion.These problems have seriously threatened people’s health and quality of life.Intelligent vehicle have the ability to avoid traffic accidents and improve the traffic environment.Therefore,driverless technology has become a social hotspot nowadays.The driverless system of the intelligent vehicle includes four important technologies: environmental perception,navigation and location,decision and planning,as well as control system.This paper focus on velocity planning and lateral-longitudinal control method for intelligent vehicle.The specific contents are as follows:Firstly,for the driving environment which is inconvenient to use the following model to achieve velocity planning,a constrained square wave velocity planning method based on fuzzy algorithm has been proposed.the curvature of the road ahead,the current lateral tracking error of the vehicle,and the friction coefficient of the road are considered as the key factors parameters which could affect the result of speed planning.the problem that the conventional speed planning method is hard to think humanized is solved by this method.The speed curve can provide a better experience for passenger with any speed controller which could stably track the target curve,and a re-planning method considering vehicle dynamics constraints is given.Secondly,in order to improve the speed control accuracy of intelligent vehicle,the method of longitudinal speed tracking controller based on model predictive control is given.The throttle/brake switching logic and the lower drive/braking dynamic model and the corresponding inverse longitudinal dynamics model are given for cooperating with the upper model predictive controller.The result of simulation on the platform of CarSim and Matlab/Simulink has showed the effectiveness of the method,the controlled vehicle can track different reference curves stably and quickly,the control of the throttle and brake is very smooth in the ideal speed curve tracking experiment.Then,the path tracking method of intelligent vehicle based on model-free adaptive predictive control is introduced for the purpose of solving the difficulty and inaccuracy of nonlinear system modeling.The content in this paper include preview heading tracking strategy for MFAPC,the design method for MFAPC controllers,and a reset mechanism for algorithm parameters for intelligent vehicle working environment.The tracking experiment of the reference path is realized by CarSim-Simulink simulation.The results show that the proposed method has strong robustness to vehicle speed and ground friction coefficient,and a better overall performance is displayed in the comparison with the MPC method.Finally,six experiments of real vehicle controlling were carried out under different reference paths and reference speeds.The experimental results show that the proposed method can complete the objectives that controlling the vehicle drive along the target path and speed.