| In recent years,with the rapid development of residents’ economic level and automobile industry technology,automobiles have entered thousands of households,and the number of residents’ automobiles has continued to rise,which has brought severe challenges to road traffic safety.In order to reduce the accident rate caused by human factors,intelligent driving technology has become an inevitable trend for future development.This thesis mainly studies lane line detection and path planning algorithms.The main research contents of this subject are as follows:(1)Carry out image preprocessing research on the pictures captured by the vehicle camera,highlight the characteristics of the lane lines,and prepare for the subsequent lane line fitting.In view of the phenomenon that the edge of the lane line is not prominent by using the common grayscale processing method,this thesis proposes a method of converting the RGB images to the Lab color gamut based on the color features,and the L and b channels are reserved to achieve the purpose of highlighting the characteristics of the lane line.For the ordinary Sobel operator,two diagonal directions are added for edge extraction to distinguish the lane lines of the road image.(2)The variable model of the combination of straight line and parabola is used to fit the lane line.For the road scene where the lane line is a straight line,this thesis uses Hough transform to fit the lane line.In order to improve the real-time performance of the algorithm,the traditional Hough transform method is improved,which reduces the amount of calculation in the processing process and the requirements for hardware equipment.For the scene of curve lane line,the pretreated road image is inversely perspective transformed to obtain a bird’s-eye view,the sliding window is used to extract the pixel point coordinates of the lane line,and the parabolic model is used to fit the curve lane line.On the basis of accurately identifying the boundary of the lane line and adapting to the curvature change of the lane line,this method can also meet the real-time requirements of the program.(3)Based on the artificial potential field method,this thesis studies the local path planning algorithm.Considering the influence of the road boundary repulsion field,a simulation experiment was carried out.The problems of target inaccessibility and local minimum in the traditional artificial potential field method are analyzed and studied.The problems are solved by introducing the influence of the distance between the vehicle and the target point into the repulsive potential field function and the simulated annealing algorithm.In view of the fact that the traditional artificial potential field method does not consider the influence of vehicle dynamics,a two-degree of freedom vehicle model is established by taking the vehicle dynamics constraints into account.The influence of obstacle speed and acceleration is introduced to make the planned path more reasonable.(4)Based on the research in this thesis,the programming is implemented.The programming simulation of the lane line preprocessing algorithm and the lane line fitting algorithm has been carried out to verify the effectiveness of the method used in this thesis,that is,the lane line boundary can be obtained clearly.MATLAB software is used to simulate the path planning of the improved artificial potential field method.The simulation results show that the method can complete the task of path planning from the vehicle position to the target point.The joint simulation platform of Matlab/Simulink software and Car Sim software is built,,and the vehicle dynamics factors are added.The path planning is carried out in the presence of dynamic obstacles,and the planned path is tracked.The simulation results confirm effectiveness and rationality of the improved artificial potential field method for local path planning. |
PDF Full Text Request