Research On Trajectory Tracking Control Algorithm Of Intelligent Vehicle

With the increase of vehicle ownership year by year,various types of traffic and environmental problems are brought about.Intelligent driving vehicles and related technical research are considered to be effective approaches to solve traffic problems.At the same time,the development of intelligent driving vehicles has the significant potential to build future urban smart transportation systems.Research on intelligent driving vehicles mainly involves the following three parts: environment perception,trajectory planning and decision-making,and trajectory tracking control.The intelligent vehicle firstly interacts with the driving scene through onboard sensors;and the desired driving trajectory is optimized baseing on the processing results of the environmental information,and then utilizing the vehicle’s underlying dynamics control to achieve the purpose of following the desired trajectory.Among them,the lateral motion control of intelligent driving vehicles is the basis for the safe and stable of autonomous driving vehicles,and is the research direction of related technologies of intelligent driving vehicles in recent years.In this paper,the control algorithm is designed by using preview theory and model prediction theory respectively in this direction,and the simulation results are verified on the experimental platform independently developed by this research group.The trajectory tracking control method based on the optimal preview theory first uses the information of the expected trajectory of the vehicle to establish the functional relationship between the steering angle of the front wheels and the lateral position deviation,and sequentially design the feedforward multi-points preview controller.In order to make full use of the lateral deviation,a sliding mode controller is added in the feedback procedure,and the stability of the controller is analyzed to explain the rationality of the sliding mode surface design.The design of the feedforward + sliding mode trajectory tracking controller is completed.Implementing verification through simulation experiments,the simulation results show that under the designed simulation conditions,the trajectory tracking performance of the feedforward + sliding mode controller is effective,and the control algorithm has good stability.In order to reduce the computational complexity of the lateral control of the intelligent driving vehicle and improve the real-time performance of the control method of the model predictive control(MPC)theoretical design trajectory tracking controller,this paper adopts the linear time-varying model predictive control method to control the lateral direction of the intelligent driving vehicle.The device is designed to add dynamics-related constraints to ensure that the derivation results of the linear time-varying model predictive control algorithm can well meet the vehicle dynamics performance.In order to ensure the stability and feasibility of the vehicle in the lateral control process,this paper constructs the constraints of the vehicle dynamic characteristics and the vehicle steering system constraints on the controller state and output during the lateral control.Experimental simulation results show that the tracking effect of the controller is good,and the corresponding curve changes smoothly.Finally,the two controllers are compared.The simulation results of the low-speed operating conditions show that the tracking effects of the two controllers are approximately the same,and the model predictive controller is more stable in the lateral position deviation curve under high-speed operating conditions.As the vehicle speed increases,the performance of the two controllers changes,the model prediction controller’s tracking effect in the straight line section is better,and the lateral position deviation at the curve all increases.The feedforward + sliding mode controller is relatively more stable and the error is smaller.Comparing the experimental results and experimental conditions of the two controllers to consider comprehensively,find a suitable control algorithm for the actual platform verification,and finally verify the actual platform for the trajectory tracking model prediction controller.The results show that the platform can complete the trajectory tracking the control command to meet the requirements of smart car trajectory following control.