Research On Path Planning Of Electric Vehicle Automatic Charging Device Based On Visual Positionin

Charging piles are the basic charging facilities for electric vehicles,and with the increase in the number of electric vehicles,they have seen rapid development.Currently,the charging devices for electric vehicles are mostly fixed manual charging piles,with a low level of automation.Secondly,a single parking space is equipped with a charging pile,which occupies a certain amount of space and is not fully utilized.In response to the above issues,this paper proposes a design scheme for an automatic charging device for electric vehicles based on visual positioning,which can achieve automatic charging for multiple vehicles,solving the problems of low automation level of existing charging equipment and occupying large parking space.The main research contents of this article are as follows.(1)According to charging requirements and application scenarios,a design scheme for the overall structure of an electric vehicle automatic charging device is proposed.The structural scheme of the automatic charging device is a five degree of freedom mechanical configuration,which is designed for the type of joints.Important structures and components are selected and processed,and an automatic charging device prototype is built.This prototype serves as the structural basis for the dual identification and positioning scheme of electric vehicles and the executive mechanism for subsequent path planning research.(2)A deep learning based dual recognition and positioning scheme for electric vehicles is proposed based on the structural characteristics of automatic charging devices.This scheme includes a global recognition and localization scheme and a local recognition and localization scheme.In the global recognition and positioning scheme,the images of electric vehicles with different parking angles and orientations are collected to create a data set.The YOLOv4 target detection algorithm is used for target recognition.The double semantic segmentation algorithm is used to extract the vehicle contour.The minimum bounding rectangle is used to obtain the spatial location information of electric vehicles,and then the location information of charging ports is converted according to the vehicle information.In the local recognition and positioning scheme,collect images of the charging port to create a dataset.Through a series of image processing processes such as extracting regions of interest from the charging port images,Hough circle detection,feature matching,and homography matrix pose solving,the precise position of the charging port is obtained.The positioning error of the charging port shall not exceed 3mm,providing accurate charging port location information for subsequent automatic charging device path planning.(3)The kinematics and dynamics model of the electric vehicle automatic charging device is established with the help of the robot toolbox.The improved D-H parameter method is used to establish the coordinate system,calculate the coordinate transformation relationship of adjacent joints,and deduce the forward and inverse kinematics formulas.Use Monte Carlo method to obtain the motion space of the device,providing support for subsequent trajectory planning.Simultaneously perform dynamic analysis,derive joint torque expression using Lagrange equation,and verify motor torque matching.(4)Compare the length of obstacle avoidance paths between RRT algorithm and RRT *in two-dimensional and three-dimensional spaces,and select RRT * algorithm for obstacle avoidance path planning.The motion trajectory of the electric vehicle automatic charging device is planned using a 3-5-3 mixed polynomial interpolation.During the charging process,considering motion constraints,a time optimization method for the motion trajectory of an automatic charging device based on an improved particle swarm optimization algorithm is proposed for the initial position of the motion and the target position given by visual positioning,to optimize the time of the planned charging path.Build a platform for optimizing the motion trajectory time of electric vehicle automatic charging devices,conduct20 charging docking experiments,and record the charging time,further proving the effectiveness of the improved particle swarm optimization algorithm based method for optimizing the motion trajectory time of automatic charging devices.