Research On Key Technologies Of Automatic Parallel Parking

With the rapid development of the automotive industry,the vehicle retention in China has sharply increased,but the number of parking spaces cannot match the vehicle retention,dramatically increasing the complexity of parking.Parking accidents occur frequently,causing huge property losses and life safety threats.Parking problem have become a major driving problem today.In order to solve the parking problem and achieve intelligent,safe,and efficient automatic parking of cars,this article first analyzes the current application status of automatic parking,and then conducts research and analysis from three aspects of parking space detection,trajectory planning,and tracking control methods.It analyzes the existing shortcomings of key technologies of automatic horizontal parking,and finally proposes the research method of this article:(1)In terms of parking space detection,Li DAR is used as the detection sensor.Firstly,a Li DAR parking space detection model is built based on Prescan and Matlab.Then,point cloud data of Li DAR is obtained through simulation experiments.The point cloud data is processed through point cloud coordinate conversion,invalid point filtering,point cloud projection,rasterization,and obstacle recognition to complete parking space information extraction;(2)In the aspect of path planning,two segments of third-order Bezier curve interpolation method is adopted,design parallel parking paths based on normal reversing methods,then,according to the characteristics of the third-order Bezier curve and the vehicle collision avoidance constraints,the control point equation of the third-order Bezier curve of the parallel parking path is designed and derived.Finally,the vehicle parameters and parking space information are substituted into the calculation to obtain a safe,collision free and curvature continuous parallel parking reference track;(3)In terms of path tracking,model predictive control algorithm is used as the trajectory tracking control algorithm.Based on model predictive control theory,a mathematical model of model predictive control is derived.Then,a model predictive controller model is built in Matlab.Finally,the feasibility of the constructed model controller is verified through joint simulation using Car Sim and Matlab.Finally,conduct a joint simulation experiment of automatic parallel parking based on Car Sim and Matlab,and the results showed that the driving trajectory matched the reference trajectory well;In order to further verify the feasibility of this method,real vehicle experiments were conducted using the 2016 model EV160 modified by BAIC New Energy as the test platform.The results showed that automatic parallel parking could be completed safely and without collisions,proving the feasibility of this method.