Local Trajectory Planning And Tracking Control Of Intelligent Vehicle Based On UniTire Model

Trajectory tracking and trajectory planning are important research areas of vehicle intelligence.The control target is four-wheel independent drive electric vehicles.The vehicle front wheel angles and the driving/braking torque of the wheels can be controlled directly.Combined with multi-degree-of-freedom vehicle models,complex tire models and model predictive control theory,local trajectory planning based on overtaking behavior in highspeed scenarios and the trajectory tracking are realized.The specific content includes the following parts.For local trajectory planning and trajectory tracking control,a single-track linear vehicle dynamics model,a nonlinear seven-degree-of-freedom vehicle dynamics model and UniTire model are established respectively.Carsim vehicle parameter modeling was carried out,and the whole vehicle model was verified according to the vehicle experimental data of field test.The basic theory of UniTire model is introduced.According to the test data of tire mechanical characteristics,the steady UniTire model formula with high accuracy in combined working conditions is identified.The influence of fitting parameters on tire mechanical properties is analyzed.The identification accuracy is compared with the Pacejka 5.2 tire model and the Similarity tire model.The MPC-LQR layered trajectory tracking controller is established based on the sevendegree-of-freedom vehicle model.Theoretical derivations of LTV-MPC and LQR are carried out.The longitudinal and lateral motion of the seven-degree-of-freedom vehicle model is decoupled.A layered trajectory tracking controller is designed.The lateral control uses model predictive control to obtain the front wheel angle and total yaw moment of the vehicle.Soft constraints are imposed on the side slip angle.Considering that the complexity of the tire model increases the difficulty of the solution,an approximate numerical method is adopted to solve the Jacobian matrix to reduce the computational complexity.The longitudinal control uses linear quadratic optimal control to obtain the total longitudinal force.The constraints of tire longitudinal force and road conditions are considered.The distribution controller determines the driving/braking torque of the four wheels according to the total yaw moment and the total longitudinal force.In the simulation experiment,the trajectory tracking results of different controllers and different tire models are compared,and the influence of the tire cornering angle constraint on the tracking effect is discussed.Based on the artificial potential field and linear single-track vehicle model,a local trajectory planner is established by nonlinear model predictive control.An artificial potential field is established according to the high-speed dynamic driving environment.The potential field is classified according to the types of obstacles.And two types of roads are considered,straight roads and curved roads.The linear monorail vehicle model is decoupled from longitudinal and lateral motion,and the artificial potential field and nonlinear MPC control are combined to establish a local trajectory planning algorithm.The lateral motion planning imposes tire slip angle constraints to obtain the driving trajectory points.The longitudinal motion planning considers the centroid acceleration constraint under the tire coupling characteristics to obtain the longitudinal speed of the vehicle.The vehicle dynamics constraints are considered in the trajectory planning to ensure the feasibility of the trajectory.The secant slop is solved in the cornering characteristics of the UniTire model online.By replacing the cornering stiffness in the linear tire model with the secant slope,the accuracy of the model and the adaptability under large side slip angle are improved.Polynomial fitting is adopted to fit the trajectory points into the planned trajectory.In the simulation experiment,a variety of traffic scenarios under different road frictional coefficients are designed,which are used to verify the algorithm of the local trajectory planning and trajectory tracking algorithm in this paper.