Research On Automatic Alignment Method Of Wireless Charging Car Based On Machine Vision

Because the battery capacity of electric vehicles is limited,and the current charging facilities are not yet complete,the charging problem of the magnetic coupling mechanism restricts the development prospects of electric vehicles.One of the hot issues in this field is to increase the mutual inductance of the coil and maximize the efficiency.Even the magnetic coupling mechanism is precisely aligned.However,in practical applications,there is often a situation in which the charging mechanism wastes charging resources due to misalignment of the coupling mechanism.Based on the background of this project,this paper has designed a registration method based on the visual method,which is of great value for promoting the industrial growth of wireless charging systems and the promotion and application of new energy vehicles.This topic is derived from the project of "Precision Positioning and Testing Technology of Electric Vehicle Wireless Charging System Based on Intelligent Perception" of State Grid.In this paper,the research status of vehicle vision system,posture detection algorithm and electric vehicle wireless charging alignment method are investigated respectively,and finally the wireless charging alignment method based on monocular machine vision and panoramic machine vision is determined.This paper designs and implements an automatic alignment method for wireless charging of electric vehicles based on monocular stereo vision based on the camera imaging model and the basic principles of machine vision.The basic principles and algorithms of monocular camera calibration are studied.A feature point region segmentation method is developed to automatically segment the target and detect feature points.Various principles and algorithms for PNP(Perspective-N-Point)pose detection are introduced.On this basis,the ground is completed Relative position and attitude detection of the transmitter and the vehicle receiver.The vision-based image processing algorithm design implements a wireless charging alignment method based on panoramic vision.Median filtering is performed on the panoramic image to filter out irrelevant information.The basic principle of the Canny operator is introduced and the contour of the parking space marking line is extracted based on this method.Next,the panoramic image was subjected to Hough transform and LSD(a Line Segment Detector)algorithm for straight line feature detection,the parking space marking line was extracted,and finally the parking space area determination method was designed,which can accurately segment the parking space area and give the parking space corner information.Finally,after market research,the hardware devices used were selected for the wireless charging alignment method based on monocular stereo vision and the panoramic vision alignment method,and then the software was written according to the design method and principle to achieve the alignment function.The experimental error analysis is carried out to verify the feasibility and robustness of the positioning of the electric vehicle wireless charging device through the machine vision scheme.