Automatic Parking System Based On Ultrasonic Sensor

Automatic parking system is an important part in the field of unmanned vehicle research.Its path planning and tracking method is the direct and key link to reflect parking intelligence level.In this paper,the four-wheel vehicle with wheel steering is taken as the research object,the fourth order b-spline theory is used to plan the parking path,and a path tracking method combining fuzzy control and kalman filter is designed to realize the accurate tracking of parking path.The main work is summarized in the following aspects.Firstly,the performance parameters of various distance sensor is compared,thefunctional requirements and extensibilityof the system areconsidered comprehensively,the ultrasonic sensoris selected as the ranging sensor of the parking environment information acquisition unit,the method of automatic detection of parking spaces based on the principle of ultrasonic ranging is designed to provide the research foundation for the following methods of automatic parallel parking path planning and tracking.Secondly,in view of the two circular arc tangent parking path of spin problems in tangent place,based on the establishment of vehicle kinematics model,the determination of minimum parking space and the analysis of obstacle avoidance constraints,a parking path planning method based on b-spline theory is proposed to make the intersection as smooth as possible to avoid vehicle turning in place and improve parking efficiency,The Matlab/Simulink model is built,and the simulation is carried out with different initial reversing positions to determine the effective area of the initial parking position under the proposed method,so as to verify the feasibility of the proposed method.Then,a path tracking method combining fuzzy control and Kalman filtering is proposed to solve the problem of high error and low accuracy of parking path tracking.In order to improve the accuracy of the output of the front wheel angle and the tracking precision of parking path,the Kalman filtering method is introduced into the fuzzy control to reduce the position deviation and heading angle deviation of the vehicle.The Matlab/Simulink model is built,and the fuzzy control method is used as the control to carry out the simulation comparison experiment.Finally,the simulation experiment is carried out to verify.The steering gear steering smart car is used as the experimental carrier,and a complete design scheme of hard and software for automatic parking is proposed.The automatic parking spot detection,parking path planning and tracking methods are converted into C language program that can be processed by the control chip to conduct the car experiment.The results show that the parking success rate is high and the system reliability is strong,the method of parking path planning and tracking is proved to be feasible.