Research On Automatic Parking Path Planning And Tracking Control In Narrow Parking Spaces

With the frequent occurrence of parking difficulties in big cities,consumers increasingly need automatic parking systems,which can solve the problem of driver parking difficulties,ensure safety and improve efficiency.This thesis studies the path planning technology and tracking control technology of automatic parking,proposes a parallel parking path planning method and designs a parking path tracking controller based on a nonlinear model predictive control method.The above is verified by simulation.Research reliability.The parking problem needs to consider car motion characteristics first,so this thesis simplified the four-wheel Ackerman steering model to a two-wheel bicycle model,derived the vehicle kinematics model at low speeds,and calculated the key points of the body relative to the center of the car’s rear axle.The point’s trajectory motion equation.In addition to car kinematics,vehicle parameter constraints and obstacle avoidance constraints also need to be considered.Therefore,this thesis analyzes the vehicle mechanical constraints such as the front wheel angle rate and the size of the front wheel angle,and specifies the curvature constraints for the planned path in this thesis.The analysis is based on circular arcs.Obstacle avoidance constraints under the parking path of the straight-line tangent method.In order to solve the problems of the parking system facing narrow parallel parking spaces that cannot find the parking path,the path curvature changes,the path does not conform to the automobile mechanical constraints and the obstacle avoidance constraints,this thesis decouples the path planning problem into the path planning of the entry section and the path planning of the adjustment section.,And take the minimum time of the entire parking process as the optimization goal.For the adjustment outside the parking space,in order to make the path curvature continuous,and to consider the obstacle avoidance constraints and the car’s own performance constraints,this thesis designs a loop curve-arc-straight line combination method to enter the parking path.For narrow parking spaces,the adjustment within the parking space pays more attention to whether the vehicle can maintain the correct posture and the front wheel angle to stop at the target end point.Therefore,this thesis designs a straight-arc combination method to adjust the parking path,and park the car in a straight line.The path and the double-arc backward parking path make the car approach the end point successively.This thesis simulates multiple initial poses in MATLAB,and the simulation results show the feasibility and strong adaptability of the method in this thesis.In view of the large curvature of the automatic parking process and the limitation of the speed and size of the front wheel angle of the car restricting the tracking control accuracy,this thesis proposes a nonlinear model predictive control method based on particle swarm optimization,and uses the model predictive control method to predict the system The future output will be optimized by particle swarm optimization in the rolling optimization part.In Simulink,this thesis simulates multiple initial poses,and the simulation results show that the method in this thesis can track the reference path well on curves with larger curvatures.Compared with NMPC,the tracking accuracy is improved.