| Today,when intelligence has become an important development strategy for all walks of life,vehicle intelligence has also become a hot research direction in the automotive industry.As one of the highlights of low-speed automatic driving in the development of vehicle intelligence,automatic parking technology,has naturally become a research hotspot for major OEMs,intelligent driving technology suppliers and universities on intelligent driving;at the same time,as the problem of urban parking becomes more and more prominent,car owners have an increasingly strong demand for parking systems,which also Accelerates the development of automatic parking technology.The automatic parking system consists of three major modules: parking space detection,path planning and path tracking control.In this thesis,the parameters of a certain car are used as the carrier to research and analyze these three parts,and focus on the path planning and path tracking of the parking system.Controls were carried out for research analysis.In terms of parking space detection,this thesis first analyzes and compares the advantages and disadvantages of various common sensors on the market in the parking space detection of automatic parking systems,and introduces the working principle of parking space detection sensors in automatic parking systems based on ultrasonic radar,and analyzes ultrasonic waves.The shortcomings of radar in the parking space detection module,secondly,the parking space detection method of fusion camera is introduced,and it has been verified by real vehicles.At the same time,according to the Ackerman steering principle,a vehicle model based on kinematics is established.In terms of path planning,this thesis adopts the method of reverse driving.First,it analyzes the preconditions of path planning,including calculating the position of the vehicle body at each moment,and obtaining the coordinates of each key point of the vehicle body.Based on the geometric method,respectively the paths of one-step parallel parking and two-step parallel parking are planned,and the minimum parking space size required by the one-step parallel parking path planning method is analyzed and calculated;for the vertical parking path planning,this thesis also The single-step vertical parking path planning is designed based on the method of geometry.At the same time,the optimized single-step vertical parking path planning is designed according to the vertical parking conditions and the different starting positions of the vehicle.Finally,the simulation analysis of parallel and vertical parking conditions is carried out using Matlab,which verifies the feasibility of the planning method.In terms of path tracking control,this thesis first introduces the model predictive control algorithm,then designs the path tracking controller of the parking system based on the model predictive control algorithm.Finally,based on the co-simulation of carsim and simulink,the feasibility of the tracking controller designed based on the model predictive control algorithm is verified. |
