Research On Dynamic Planning And Trajectory Tracking Control Of Autonomous Vehicles

In recent years,with the rapid development of computer technology and sensor technology,driverless technology has gradually become a research hotspot in various scientific research institutions and universities.Dynamic trajectory planning and motion control,as an important part of the unmanned system,directly affect the performance of the unmanned system and play a pivotal role.Commonly used dynamic trajectory planning algorithms generally use geometric curves to smooth and optimize the planned trajectory,and set certain geometric constraints in the process to make the trajectory meet the kinematic constraints of vehicle driving.However,the kinematics of the vehicle change with the driving conditions of the vehicle,and simply using geometric constraints cannot be applied to all driving conditions.In view of the above problems,this paper uses the kinematics model of the vehicle to design a trajectory planner based on model predictive control,and introduces the obstacle avoidance function to make the planner have the ability to avoid obstacles.When the vehicle is driving in high-speed scenarios,the dynamic characteristics of the vehicle will become more complex,and accordingly,the nonlinear constraints of the vehicle will become more stringent.Therefore,the traditional control algorithm designed based on motion control is difficult to meet the control requirements of the vehicle at high speed.Aiming at the above problems,this paper uses the vehicle dynamics model to design a trajectory tracking lateral controller based on model predictive control to improve the control stability of the trajectory tracking controller under different driving conditions.The specific work includes the following aspects:（1）By establishing the vehicle kinematics model in the Frenet coordinate system,the longitudinal planning and lateral planning of the vehicle are decoupled,and the planning task is simplified.Based on the model predictive control algorithm,a local dynamic trajectory planner is designed to improve the adaptability of the planner under different driving conditions.The traditional obstacle avoidance function is improved,and the shape information of the obstacle is introduced to improve the adaptability of the planner to obstacles of different sizes.（2）Aiming at the working conditions of the vehicle at high speed,a dynamic model of the vehicle is established based on the dynamic characteristics of the vehicle,which improves the stability of the control algorithm under different driving conditions,and uses coordinate transformation to transform the dynamics of the vehicle.The model is transformed to the Frenet coordinate system,which decouples the lateral and longitudinal control of the vehicle and simplifies the design of the c ontroller.The dynamic model is linearized and discretized at a small angle,and a quadratic objective function is established according to the dynamic constraints,and the QP solver is used to solve the problem to improve the calculation speed of the controller.A displacement-velocity closed-loop PID controller is designed for longitudinal control of the vehicle.（3）Deploy the local dynamic trajectory planner and trajectory tracking controller to the real vehicle,and verify the planning effect and control effect of the algorithm on the real vehicle through real vehicle testing.