Research On Path Planning And Tracking Control Of Automatic Parking System Based On MPC

The rapid development of science and technology has brought great influence to the automotive industry,and the intelligent network connected vehicles have become a new growth pole in the automotive industry.In this paper,the path decision planning and path tracking control algorithm of automatic parking system are analyzed and studied in order to improve the safety and reliability of parking process.The main research work is as follows:Firstly,ultrasonic wave,laser rangefinder and wheel speed sensor are used as sensing components of parking space detection,and the principle of parking space detection under parallel and vertical parking conditions is introduced.Aiming at the problem that the ultrasonic radar beam angle causes the detection error,the parking space detected by the radar is corrected by the parking space compensation method and the beam angle calibration experiment is carried out by using the ultrasonic ranging module.Two remote ultrasonic sensors on the same side of the vehicle and the laser rangefinder on the front wheel are used for detection successively.Combined with the data measured by the three sensors,the final size of the parking space can be obtained,so as to improve the detection accuracy and prevent the failure of a single sensor from affecting the normal operation of the system.Secondly,the vehicle kinematics model based on Ackerman steering principle is established,and the correctness of the established model is verified by simulation experiments.The parking system coordinates are analyzed,and the minimum parking space size is preliminarily calculated in the case of single step parking.the parking path is preliminarily planned in the case of arc line combination path,and the starting point area of two parking conditions is calculated,which lays the foundation for the subsequent parking path planning.Thirdly,the constraints in the process of path planning are analyzed,including environmental constraints,vehicle kinematics constraints and vehicle pose constraints.The path curve of high-order polynomial planning selected by parallel and vertical parking conditions is simulated by MATLAB to verify the feasibility of the path planning method.Finally,a Simulink/Car Sim joint simulation platform was built to compare the performance of the path tracking preview control algorithm based on pure pursuit and the MPC path tracking controller based on incremental model.The simulation results show that the effect of parking path tracking controller based on MPC is better than that of path tracking preview control based on pure pursuit algorithm.The control model code is loaded into mini Intel I3 industrial computer to control ROS intelligent car for parking experiment,which verifies the feasibility of the designed parking controller.