Trajectory Tracking And Path Planning Research For Driverless Cars

With the rapid development of science and technology for decades,autonomous driving technology has promoted the progress of the entire society,affecting the way humans travel,and has received a lot of attention from all walks of life.At present,the core of driverless technology mainly includes environmental awareness,path planning and trajectory tracking.The research focus of this paper will be on the path planning and trajectory tracking of smart cars.Firstly,aiming at the path tracking control problem of unmanned vehicles,a lateral tracking controller based on model prediction control algorithm was designed,a three-degree-of-freedom vehicle dynamics model was established,the objective function was integrated,and constraints were added.Under the joint Carsim/Simulink simulation platform,the tracking of the MPC controller when the vehicle turns at different speeds was analyzed.In addition,the fuzzy PID controller was used to compare,thereby determining that the MPC controller has higher tracking accuracy,better stability and higher robust performance.Then,an improved Bi-RRT algorithm was proposed on the path planning problem,using the target gravitational idea in the artificial potential field method,and adding the target gravitational function on the basis of the Bi-RRT algorithm.The redundant nodes were eliminated by the distance threshold method,and then the three-dimensional average B-spline interpolation function was added to smooth the path,so that the planned path is more reasonable and smooth,meeting the requirements of vehicle steering.In addition,according to the known map information,the improved algorithm,RRT algorithm,and Bi-RRT algorithm were simulated in three different environments.The conclusion is that the improved Bi-RRT algorithm can greatly improve the search efficiency and shorten the running time.Finally,in order to verify the automatic navigation performance of unmanned vehicles in the actual driving process,a real car experimental platform was built,the hardware and software system of the vehicle were introduced,and the path planning and tracking control of the real car were tested.Experimental results show that the improved Bi-RRT algorithm can meet the fast search of global paths,and the completion time is relatively short.The MPC landscape controller also has good tracking effect.