Electric Vehicle Charging Robot Charging Port Identification Method Based on Multi-algorithm Fusion

Abstract

Aiming at the problem of plugging and positioning identification in the process of automatic charging of electric vehicles, especially the difficult problem of low efficiency and accuracy of identification in complex operation environment, this article proposes a multi-algorithm fusion of electric vehicle charging port identification method, which can effectively obtain the characteristic information of the round hole of charging port and realize the purpose of automatic identification of charging port by robot. Firstly, an electric vehicle charging port in a complex environment is analyzed and simulated, and camera selection and calibration are explained; on this basis, algorithms based on image smoothing filtering, feature detection segmentation of ROI region, improved Canny edge detection and combined mathematical morphology are proposed to correlate the charging port image respectively, and the features of the target charging port jack are extracted. Finally, the experimental verification of the charging port identification method was conducted for different illumination intensities and different shooting distances. The experimental results show that the identification success rate is 93.3% under the weak illumination and 97.8% under the normal illumination intensity of 4000lx. This shows that the method can effectively improve the robustness and accuracy of the charging port identification.