Research On Driving State Estimation And Trajectory Tracking Control Of Commercial Vehicles

Trajectory tracking control is a key component of autonomous driving technology.Commercial vehicles can be used in special usage scenarios,and it is of great prospect for autonomous driving technology in reducing driver burden and reducing traffic accidents.Hence,this paper takes commercial vehicles as the research object,considering the high center of gravity and large changes in vehicle parameters before and after loading of commercial vehicles.It designs a vehicle driving state estimator and trajectory tracking controller to achieve real-time acquisition of vehicle driving state information and lane keeping and following,where the state estimation model estimates unknown quantities based on existing vehicle state information.As the input of the trajectory tracking control model,the following of the desired trajectory is realized.First,based on the characteristics of road driving for commercial vehicles,a nonlinear dynamic model for the target vehicle is established,which includes longitudinal,lateral,and yaw dynamics of the vehicle.A tire mechanics model is also established,and simulation conditions are set up to verify the dynamic model of the vehicle.Second,considering that the vehicle system has strong nonlinearity when driving under conditions of large curvature radius and low road adhesion coefficient,traditional state estimation based on extended Kalman filter algorithm suffers from truncation error,resulting in large deviations between the estimated value and the actual value.To avoid this issue,a state estimation model based on unscented Kalman filter algorithm is designed to obtain the vehicle driving state information.In addition,hypothetical estimation is carried out for the loaded vehicle parameters of the commercial vehicle,and simulation is designed to verify the accuracy of the state estimation model.The simulation results show that the vehicle state estimation model can estimate the unknown parameters based on known quantities.In addition,a vehicle trajectory tracking model is designed to simplify the model and improve the accuracy of vehicle trajectory tracking.The longitudinal and lateral movements are decoupled and controlled separately.The PID algorithm is used to control the longitudinal velocity of the vehicle to achieve the desired speed following.The lateral control strategy adopts the LQR control algorithm,and an error model is established based on the actual driving information of the vehicle and the planned trajectory information.To counteract the error interference,a feedforward control quantity is introduced into the objective function to improve the accuracy of the algorithm,and simulation conditions are set to verify the accuracy of the vehicle tracking model.Finally,a simulation testing environment is built based on MATLAB/Truck Sim software to verify the effectiveness of the vehicle state estimation and trajectory tracking under different road conditions.Simulation results show that:1.Under high-road-adhesion conditions and variable-speed conditions,both the UKF algorithm-based and EKF algorithm-based state estimation models exhibit high accuracy,and the trajectory tracking control model can follow the desired trajectory.This is because the road surface can provide sufficient tire force under such conditions.2.Under low-road-adhesion conditions and variable-speed conditions,the nonlinear characteristics of the vehicle become stronger,resulting in a decrease in the accuracy of the estimation and control models.However,the UKF estimation model is superior to the traditional EKF algorithm model.Although the vehicle can still follow the desired trajectory,further parameter modification is needed.Additionally,simulation analysis was performed on the adaptability of the model parameters before and after loading,and the results show that the adaptability of the vehicle parameters has a significant impact on the trajectory tracking accuracy.