Research On Longitudinal And Lateral Coupling Control For Trajectory Tracking Of In-wheel Motor Driven Vehicles

In-wheel motor driven vehicles oriented to intelligent driving technology have broad application prospects and are the ultimate direction for the development of intelligent networked vehicles.The trajectory tracking control algorithm is the last link of the perception-planning-decision-control system in the intelligent driving domain,which determines the quality of the final execution effect.In view of the unique redundant control properties of in-wheel motor driven vehicles,it is of practical guiding significance to conduct research on trajectory tracking vertical and horizontal comprehensive control algorithms based on vehicle driving state parameter estimation for the intelligent driving domain.For this reason,this paper takes the four-in-wheel motor driven vehicle as the research object,and the main research contents are as follows:(1)Establish the vehicle dynamics model of in-wheel motor driven vehicle and carry out simulation verification.For the dynamic model of the vehicle body,the nonlinear 7degrees of freedom(7-DOF)dynamics model is established by fully considering the independent controllable characteristics of wheel speed and torque of in-wheel motordriven vehicle.Meanwhile,the Uni Tire tire model with excellent generalization ability is built to reflect the strong nonlinear coupling characteristics of the vehicle.The simulation verification shows that the built model has good accuracy at low,medium and high speeds,and the overall error is kept within 5%;(2)Design a vehicle state parameter estimator for intelligent driving domain.Firstly,the kinematic integrator is combined with the redundant information of wheel speed and the slip rate information to estimate the longitudinal speed,and the accuracy of longitudinal speed estimation is improved.Secondly,the strong tracking fade factor is introduced into the adaptive unscented Kalman filter algorithm to perform the cascaded estimation of tire force and centroid side deflection Angle.Combined with the kinematic estimation,the adaptive weight factor based on the difference of front and rear wheel side deflection Angle is given to further optimize the estimation of centroid side deflection Angle and improve the real-time performance and robustness of the algorithm.The simulation results show that the estimation accuracy of the proposed estimation algorithm for parameters such as longitudinal velocity,yaw rate,and side slip angle is 11.65%,10.62%,and 12.80% at low speeds,and 16.16%,15.48%,and 19.41% at medium speeds,respectively.(3)Design and improvement of longitudinal and lateral trajectory tracking control algorithm considering stability control.Firstly,model predictive control(MPC)is used to calculate the total longitudinal driving demand torque for the longitudinal velocity control algorithm,and it is used as the hard constraint of the lateral trajectory tracking control layer.For lateral trajectory tracking control,in order to solve the control law based on discrete time domain,the inverse error matrix is introduced to improve the nonlinear model predictive control algorithm,and the fourth-order Runge-Kutta method is used to discretization the state variables and control variables.Secondly,the stability control strategy is designed based on the phase plane of centroid side deflection angle,and the phase plane stability boundary is determined by double-line method.The ideal steadystate model is used to calculate the quadratic programming problem to solve the additional yaw moment,and the yaw moment calculated by the longitudinal and transverse trajectory tracking control layer is corrected to keep the vehicle in the phase plane stability domain of centroid side deflection Angle.(4)Optimal allocation of driving torque based on tire utilization dynamic efficiency matrix.The elliptic method was used to fit the vertical and horizontal tire force coupling characteristic curves,construct a three-dimensional elliptic parameter table,update the tire utilization efficiency matrix in real time,and obtain the driving torque of four wheels by solving the optimization problem to realize the final executive control.(5)Vehicle test.The domain control dynamics test platform of hub motor drive is built with GNSS and IMU as the high-precision positioning and attitude solution,and the algorithm software architecture is built with ROS as the operating framework.In order to verify the effectiveness of the proposed state parameter estimation algorithm and the synthetic control algorithm,ramp entry and emergency collision avoidance driving scenarios is simulated under constant curvature and snake test conditions respectively.The experimental results show that the average absolute value errors of the proposed integrated architecture algorithm are both less than 0.1 m,showing a good effect,and the average calculation time cost is less than 15 ms,while ensuring the real-time requirement of the calculation.