Research On Automatic Parking Trajectory Planning And Tracking Control Based On Vehicle Model

With the rapid growth of car ownership,not only has the traffic congested day by day,but the parking space allocated to each car is also gradually shrinking.The small parking space increases the difficulty of parking operations,which in turn leads to a significant increase in the incidence of parking-related traffic safety accidents.Therefore,research on automatic parking technology can bring huge social benefits.The research in this paper is mainly focused on the two parts of trajectory planning and trajectory tracking in the automatic parking system.The main research contents are as follows:(1)Research on parking trajectory planning based on vehicle kinematics model.First,a vehicle kinematics model was established based on Ackerman’s steering principle,and constraints were analyzed based on the built parking scene model and vehicle structural performance,including collision constraints,steering performance constraints,and geometric constraints.Secondly,this paper sets the starting point of the parking trajectory planning as the parking position of the vehicle after parking in the parking space,planning from the inside of the high-constrained parking space to the outside of the low-constrained parking space,by analyzing the turning angle and speed of the vehicle during the parking movement.The change trend,after simplified processing,designed a turning angle and speed change curve that includes all feasible parking trajectories in a simplified state,and a series of parking trajectory clusters that move according to the designed turning angle and speed curve are obtained through simulation,and combined with constraint conditions,Designed the parking trajectory optimization function,and finally obtained the optimal parking trajectory that satisfies the constraints and is easy for the vehicle to track from the feasible parking trajectory cluster.(2)Research on model predictive trajectory tracking controller based on three-degree-of-freedom dynamic model.Model predictive control algorithms are widely used because they can add various vehicle dynamics constraints in the controller process,and can improve control accuracy and robustness.This paper first established a three-degree-of-freedom dynamic model including lateral,longitudinal and yaw motions,and used it as a predictive model to design a model to predict the lateral turning angle tracking controller,which was established according to the expected trajectory parameters,the initial state of the vehicle,and the real-time state of the vehicle The objective function is to obtain the control value of the turning angle by solving the objective function;in addition,this paper designs a longitudinal speed controller based on the model predictive controller,and obtains the expected acceleration value at each moment according to the expected speed and the actual vehicle speed and acceleration value.A proportional controller calculates the vehicle’s longitudinal control amount throttle opening and brake master cylinder pressure.The horizontal and vertical controllers are jointly controlled to achieve precise tracking control of the desired trajectory.(3)Co-simulation analysis was performed on the trajectory tracking control method designed in this paper.The professional vehicle dynamics simulation software Car Sim was used to establish a simulated vehicle model close to the actual situation,and a parking driving scene reflecting the real situation was built in Car Sim.Then use Simulink to write the parking trajectory tracking control function,then design the interface between Car Sim and Simulink,and build the Car Sim/Simulink co-simulation model.The simulation results show that the model predictive trajectory tracking controller based on the three-degree-of-freedom dynamic model designed in this paper has ideal trajectory tracking accuracy and good stability.Finally,this article uses Jetson-TX2 as the controller,uses the Car Sim model to simulate a real vehicle,builds a hardware-in-the-loop simulation experiment through the CAN signal communication module of Simulink,and simulates the movement process of the real vehicle controller controlling the underlying actuator through CAN signal transmission.The hardware is in the results of loop simulation experiments show that the trajectory tracking controller designed in this paper has a good control effect and can meet the requirements of real-time and precision in actual vehicle control.